CO

COMPUTER ORGANIZATION 10CS46
COMPUTER ORGANIZATION
Subject Code: 10CS46
PART A
om
UNIT-1 6 Hours
Basic Structure of Computers: Computer Types, Functional Units, Basic Operational
c
Concepts, Bus Structures, Performance Processor Clock, Basic Performance Equation,
p.
Clock Rate, Performance Measurement, Historical Perspective
ou
Machine Instructions and Programs: Numbers, Arithmetic Operations and Characters,
Memory Location and Addresses, Memory Operations, Instructions and Instruction
Sequencing,
gr
UNIT - 2
ts 7 Hours
Machine Instructions and Programs contd.: Addressing Modes, Assembly Language,
en
Basic Input and Output Operations, Stacks and Queues, Subroutines, Additional
Instructions, Encoding of Machine Instructions
ud
UNIT - 3 6 Hours
st
Input/Output Organization: Accessing I/O Devices, Interrupts Interrupt Hardware,

Enabling and Disabling Interrupts, Handling Multiple Devices, Controlling Device
ity
Requests, Exceptions, Direct Memory Access, Buses

.c
UNIT-4 7 Hours
w
Input/Output Organization contd.: Interface Circuits, Standard I/O Interfaces PCI

Bus, SCSI Bus, USB
w
w
Dept of CSE,SJBIT Page 1

PART B
UNIT - 5 7 Hours
Memory System: Basic Concepts, Semiconductor RAM Memories, Read Only
Memories, Speed, Size, and Cost, Cache Memories Mapping Functions, Replacement
om
Algorithms, Performance Considerations, Virtual Memories, Secondary Storage
c
UNIT - 6 7 Hours
p.
Arithmetic: Addition and Subtraction of Signed Numbers, Design of Fast Adders,
Multiplication of Positive Numbers, Signed Operand Multiplication, Fast Multiplication,
ou
Integer Division, Floating-point Numbers and Operations
gr
UNIT - 7 6 Hours
Basic Processing Unit: Some Fundamental Concepts, Execution of a Complete
ts
Instruction, Multiple Bus Organization, Hard-wired Control, Microprogrammed Control
en
UNIT - 8 6 Hours
Multicores, Multiprocessors, and Clusters: Performance, The Power Wall, The Switch
ud
from Uniprocessors to Multiprocessors, Amdahls Law, Shared Memory

Multiprocessors, Clusters and other Message Passing Multiprocessors, Hardware
st
Multithreading, SISD, IMD, SIMD, SPMD, and

ity
Vector.
.c
Text Books:
1. Carl Hamacher, Zvonko Vranesic, Safwat Zaky: Computer Organization, 5th Edition,
w
Tata McGraw Hill, 2002. (Listed topics only from Chapters 1, 2, 4, 5, 6, 7)

w
2. David A. Patterson, John L. Hennessy: Computer Organization and Design The

Hardware / Software Interface ARM Edition, 4th Edition, Elsevier, 2009.
w

TABLE OF CONTENTS
PART A
om
UNIT - 1: Basic Structure of Computers Machine Instructions and Programs6-33
UNIT-2: Machine Instructions and Programs contd...34-58
c
UNIT-3: Input/output Organization.59-88
p.
UNIT-4 : Input/output Organization contd89-112
ou
PART B
gr
UNIT-5: Memory System...113-151
ts
UNIT-6: Arithmetic152-177
UNIT-7: Basic Processing Unit......178-200
en
UNIT-8: Multicores, Multiprocessors, and Clusters...201-215
ud
st
ity
.c
w
w
w

UNIT-1
Basic Structure of Computers: Computer Types, Functional Units, Basic
Operational Concepts, Bus Structures, Performance Processor Clock,
om
Basic Performance Equation, Clock Rate, Performance Measurement,
Historical Perspective.
c
Machine Instructions and Programs: Numbers, Arithmetic Operations
p.
and Characters, Memory Location and Addresses, Memory Operations,
ou
Instructions and Instruction Sequencing.
gr
ts
en
ud
st
ity
.c
w
w
w

CHAPTER 1
BASIC STRUCTURE OF COMPUTERS
1.1 Computer types
om
A computer can be defined as a fast electronic calculating machine that accepts
the (data) digitized input information process it as per the list of internally stored
instructions and produces the resulting information.
c
p.
List of instructions are called programs & internal storage is called computer
memory.
ou
The different types of computers are
1. Personal computers: - This is the most common type found in homes, schools,
gr
Business offices etc., It is the most common type of desk top computers with
processing and storage units along with various input and output devices.
ts
2. Note book computers: - These are compact and portable versions of PC
en
3. Work stations: - These have high resolution input/output (I/O) graphics
capability, but with same dimensions as that of desktop computer. These are used
ud
in engineering applications of interactive design work.

4. Enterprise systems: - These are used for business data processing in medium to
large corporations that require much more computing power and storage capacity
st
than work stations. Internet associated with servers have become a dominant
ity
worldwide source of all types of information.

5. Super computers: - These are used for large scale numerical calculations
.c
required in the applications like weather forecasting etc.,

w
1.2 Functional unit

w
A computer consists of five functionally independent main parts input, memory,

arithmetic logic unit (ALU), output and control unit.
w

Input ALU
I/O Processor
Memory
om
Output Control Unit
c
p.
ou
Fig a : Functional units of computer
Input device accepts the coded information as source program i.e. high level
language. This is either stored in the memory or immediately used by the processor to
gr
perform the desired operations. The program stored in the memory determines the
processing steps. Basically the computer converts one source program to an object
ts
program. i.e. into machine language.
en
Finally the results are sent to the outside world through output device. All of
these actions are coordinated by the control unit.
ud
Input unit: -
The source program/high level language program/coded information/simply data
is fed to a computer through input devices keyboard is a most common type. Whenever a
st
key is pressed, one corresponding word or number is translated into its equivalent binary
code over a cable & fed either to memory or processor.
ity
Joysticks, trackballs, mouse, scanners etc are other input devices.

.c
Memory unit: -
w
Its function into store programs and data. It is basically to two types
w
1. Primary memory
2. Secondary memory
w
1. Primary memory: - Is the one exclusively associated with the processor and operates
at the electronics speeds programs must be stored in this memory while they are being
executed. The memory contains a large number of semiconductors storage cells. Each

capable of storing one bit of information. These are processed in a group of fixed site
called word.
To provide easy access to a word in memory, a distinct address is associated with

each word location. Addresses are numbers that identify memory location.
om
Number of bits in each word is called word length of the computer. Programs
must reside in the memory during execution. Instructions and data can be written into the
memory or read out under the control of processor.
c
Memory in which any location can be reached in a short and fixed amount of
p.
time after specifying its address is called random-access memory (RAM).
ou
The time required to access one word in called memory access time. Memory
which is only readable by the user and contents of which cant be altered is called read
only memory (ROM) it contains operating system.
gr
Caches are the small fast RAM units, which are coupled with the processor and
ts
are aften contained on the same IC chip to achieve high performance. Although primary
storage is essential it tends to be expensive.
en
2 Secondary memory: - Is used where large amounts of data & programs have to be
stored, particularly information that is accessed infrequently.
ud
Examples: - Magnetic disks & tapes, optical disks (ie CD-ROMs), floppies etc.,
Arithmetic logic unit (ALU):-

st
Most of the computer operators are executed in ALU of the processor like
addition, subtraction, division, multiplication, etc. the operands are brought into the ALU
ity
from memory and stored in high speed storage elements called register. Then according
to the instructions the operation is performed in the required sequence.
.c
The control and the ALU are may times faster than other devices connected to a
w
computer system. This enables a single processor to control a number of external devices
such as key boards, displays, magnetic and optical disks, sensors and other mechanical
w
controllers.
w
Output unit:-
These actually are the counterparts of input unit. Its basic function is to send the
processed results to the outside world.
Examples:- Printer, speakers, monitor etc.

Control unit:-
It effectively is the nerve center that sends signals to other units and senses their
states. The actual timing signals that govern the transfer of data between input unit,
processor, memory and output unit are generated by the control unit.
om
1.3 Basic operational concepts
c
To perform a given task an appropriate program consisting of a list of instructions is
stored in the memory. Individual instructions are brought from the memory into the
p.
processor, which executes the specified operations. Data to be stored are also stored in
the memory.
ou
Examples: - Add LOCA, R0
gr
This instruction adds the operand at memory location LOCA, to operand in
register R0 & places the sum into register. This instruction requires the performance of
several steps,
ts
en
1. First the instruction is fetched from the memory into the processor.
ud
2. The operand at LOCA is fetched and added to the contents of R0

3. Finally the resulting sum is stored in the register R0
st
The preceding add instruction combines a memory access operation with an ALU
ity
Operations. In some other type of computers, these two types of operations are performed
by separate instructions for performance reasons.
.c
Load LOCA, R1
w
Add R1, R0
Transfers between the memory and the processor are started by sending the
w
address of the memory location to be accessed to the memory unit and issuing the
w
appropriate control signals. The data are then transferred to or from the memory.

MEMORY
c om
p.
MAR MDR
CONTROL
ou
PC R0
gr
R1

ts
ALU
IR

en
ud
n- GPRs
st
Fig b : Connections between the processor and the memory

ity
The fig shows how memory & the processor can be connected. In addition to the
ALU & the control circuitry, the processor contains a number of registers used for several
.c
different purposes.
w
The instruction register (IR):- Holds the instructions that is currently being executed.
Its output is available for the control circuits which generates the timing signals that
w
control the various processing elements in one execution of instruction.

w
The program counter PC:-

This is another specialized register that keeps track of execution of a program. It
contains the memory address of the next instruction to be fetched and executed.
Besides IR and PC, there are n-general purpose registers R0 through Rn-1.

Downloaded from www.citystudentsgroup.com
The other two registers which facilitate communication with memory are: -
1. MAR (Memory Address Register):- It holds the address of the location to be
accessed.
2. MDR (Memory Data Register):- It contains the data to be written into or read
out of the address location.
om
Operating steps are
1. Programs reside in the memory & usually get these through the I/P unit.
c
2. Execution of the program starts when the PC is set to point at the first instruction
of the program.
p.
3. Contents of PC are transferred to MAR and a Read Control Signal is sent to the
memory.
ou
4. After the time required to access the memory elapses, the address word is read out
of the memory and loaded into the MDR.
5. Now contents of MDR are transferred to the IR & now the instruction is ready to
gr
be decoded and executed.
6. If the instruction involves an operation by the ALU, it is necessary to obtain the
ts
required operands.
7. An operand in the memory is fetched by sending its address to MAR & Initiating
en
a read cycle.
8. When the operand has been read from the memory to the MDR, it is transferred
from MDR to the ALU.
ud
9. After one or two such repeated cycles, the ALU can perform the desired
operation.
10. If the result of this operation is to be stored in the memory, the result is sent to
st
MDR.
11. Address of location where the result is stored is sent to MAR & a write cycle is
ity
initiated.
12. The contents of PC are incremented so that PC points to the next instruction that
.c
is to be executed.
w
Normal execution of a program may be preempted (temporarily interrupted) if

some devices require urgent servicing, to do this one device raises an Interrupt signal.
w
An interrupt is a request signal from an I/O device for service by the processor.
w
The processor provides the requested service by executing an appropriate interrupt

service routine.
The Diversion may change the internal stage of the processor its state must be
saved in the memory location before interruption. When the interrupt-routine service is

completed the state of the processor is restored so that the interrupted program may
continue.
1.4 Bus structure
The simplest and most common way of interconnecting various parts of the
om
computer. To achieve a reasonable speed of operation, a computer must be organized so
that all its units can handle one full word of data at a given time.A group of lines that
serve as a connecting port for several devices is called a bus.
c
p.
In addition to the lines that carry the data, the bus must have lines for address and
control purpose. Simplest way to interconnect is to use the single bus as shown
ou
INPUT MEMORY PROCESSOR OUTPUT
gr
ts
en
ud
st
Fig c: Single bus structure

ity
Since the bus can be used for only one transfer at a time, only two units can
actively use the bus at any given time. Bus control lines are used to arbitrate multiple
requests for use of one bus.
.c
Single bus structure is

w
Low cost
w
Very flexible for attaching peripheral devices

w
Multiple bus structure certainly increases, the performance but also increases the
cost significantly.

All the interconnected devices are not of same speed & time, leads to a bit of a
problem. This is solved by using cache registers (ie buffer registers). These buffers are
electronic registers of small capacity when compared to the main memory but of
comparable speed.
The instructions from the processor at once are loaded into these buffers and then
om
the complete transfer of data at a fast rate will take place.
1.5 Performance
c
p.
The most important measure of the performance of a computer is how quickly it
can execute programs. The speed with which a computer executes program is affected by
the design of its hardware. For best performance, it is necessary to design the compiles,
ou
the machine instruction set, and the hardware in a coordinated way.
gr
The total time required to execute the program is elapsed time is a measure of the
performance of the entire computer system. It is affected by the speed of the processor,
the disk and the printer. The time needed to execute a instruction is called the processor
ts
time.
en
Just as the elapsed time for the execution of a program depends on all units in a
computer system, the processor time depends on the hardware involved in the execution
of individual machine instructions. This hardware comprises the processor and the
ud
memory which are usually connected by the bus as shown in the fig c.
st
Cache
ity
Main Processor
Memory Memory
.c
w
Bus
w
w
Fig d: The processor cache
The pertinent parts of the fig. c are repeated in fig. d which includes the cache
memory as part of the processor unit.

Let us examine the flow of program instructions and data between the memory
and the processor. At the start of execution, all program instructions and the required data
are stored in the main memory. As the execution proceeds, instructions are fetched one
by one over the bus into the processor, and a copy is placed in the cache later if the same
instruction or data item is needed a second time, it is read directly from the cache.
om
The processor and relatively small cache memory can be fabricated on a single
IC chip. The internal speed of performing the basic steps of instruction processing on
c
chip is very high and is considerably faster than the speed at which the instruction and
data can be fetched from the main memory. A program will be executed faster if the
p.
movement of instructions and data between the main memory and the processor is
minimized, which is achieved by using the cache.
ou
For example:- Suppose a number of instructions are executed repeatedly over a short
period of time as happens in a program loop. If these instructions are available in the
gr
cache, they can be fetched quickly during the period of repeated use. The same applies to
the data that are used repeatedly.
ts
Processor clock: -
en
Processor circuits are controlled by a timing signal called clock. The clock
designer the regular time intervals called clock cycles. To execute a machine instruction
the processor divides the action to be performed into a sequence of basic steps that each
ud
step can be completed in one clock cycle. The length P of one clock cycle is an important
parameter that affects the processor performance.
st
Processor used in todays personal computer and work station have a clock rates
that range from a few hundred million to over a billion cycles per second.
ity
1.6 Basic performance equation

.c
We now focus our attention on the processor time component of the total elapsed
w
time. Let T be the processor time required to execute a program that has been prepared
in some high-level language. The compiler generates a machine language object program
w
that corresponds to the source program. Assume that complete execution of the program
requires the execution of N machine cycle language instructions. The number N is the
w
actual number of instruction execution and is not necessarily equal to the number of
machine cycle instructions in the object program. Some instruction may be executed
more than once, which in the case for instructions inside a program loop others may not
be executed all, depending on the input data used.

Suppose that the average number of basic steps needed to execute one machine
cycle instruction is S, where each basic step is completed in one clock cycle. If clock rate
is R cycles per second, the program execution time is given by
N S
T=
om
R
this is often referred to as the basic performance equation.
We must emphasize that N, S & R are not independent parameters changing one
c
may affect another. Introducing a new feature in the design of a processor will lead to
p.
improved performance only if the overall result is to reduce the value of T.
ou
Pipelining and super scalar operation: -
We assume that instructions are executed one after the other. Hence the value of
S is the total number of basic steps, or clock cycles, required to execute one instruction.
gr
A substantial improvement in performance can be achieved by overlapping the execution
of successive instructions using a technique called pipelining.
ts
Consider Add R1 R2 R3
en
This adds the contents of R1 & R2 and places the sum into R3.
The contents of R1 & R2 are first transferred to the inputs of ALU. After the
ud
addition operation is performed, the sum is transferred to R3. The processor can read the
next instruction from the memory, while the addition operation is being performed. Then
of that instruction also uses, the ALU, its operand can be transferred to the ALU inputs at
st
the same time that the add instructions is being transferred to R3.
ity
In the ideal case if all instructions are overlapped to the maximum degree
possible the execution proceeds at the rate of one instruction completed in each clock
.c
cycle. Individual instructions still require several clock cycles to complete. But for the
purpose of computing T, effective value of S is 1.
w
A higher degree of concurrency can be achieved if multiple instructions pipelines

w
are implemented in the processor. This means that multiple functional units are used
creating parallel paths through which different instructions can be executed in parallel
w
with such an arrangement, it becomes possible to start the execution of several

instructions in every clock cycle. This mode of operation is called superscalar execution.
If it can be sustained for a long time during program execution the effective value of S
can be reduced to less than one. But the parallel execution must preserve logical
correctness of programs, that is the results produced must be same as those produced by

the serial execution of program instructions. Now a days may processor are designed in
this manner.
1.7 Clock rate
These are two possibilities for increasing the clock rate R.
om
1. Improving the IC technology makes logical circuit faster, which reduces the time
of execution of basic steps. This allows the clock period P, to be reduced and the
c
clock rate R to be increased.
2. Reducing the amount of processing done in one basic step also makes it possible
p.
to reduce the clock period P. however if the actions that have to be performed by
an instructions remain the same, the number of basic steps needed may increase.
ou
Increase in the value R that are entirely caused by improvements in IC
technology affects all aspects of the processors operation equally with the exception of
gr
the time it takes to access the main memory. In the presence of cache the percentage of
accesses to the main memory is small. Hence much of the performance gain excepted
ts
from the use of faster technology can be realized.
en
Instruction set CISC & RISC:-
Simple instructions require a small number of basic steps to execute. Complex
instructions involve a large number of steps. For a processor that has only simple
ud
instruction a large number of instructions may be needed to perform a given

programming task. This could lead to a large value of N and a small value of S on the
other hand if individual instructions perform more complex operations, a fewer
st
instructions will be needed, leading to a lower value of N and a larger value of S. It is not
obvious if one choice is better than the other.
ity
But complex instructions combined with pipelining (effective value of S 1)

.c
would achieve one best performance. However, it is much easier to implement efficient
pipelining in processors with simple instruction sets.
w
1.8 Performance measurements

w
It is very important to be able to access the performance of a computer, comp

designers use performance estimates to evaluate the effectiveness of new features.
w
The previous argument suggests that the performance of a computer is given by

the execution time T, for the program of interest.

Inspite of the performance equation being so simple, the evaluation of T is

highly complex. Moreover the parameters like the clock speed and various architectural
features are not reliable indicators of the expected performance.
Hence measurement of computer performance using bench mark programs is

done to make comparisons possible, standardized programs must be used.
om
The performance measure is the time taken by the computer to execute a given
bench mark. Initially some attempts were made to create artificial programs that could be
c
used as bench mark programs. But synthetic programs do not properly predict the
performance obtained when real application programs are run.
p.
A non profit organization called SPEC- system performance evaluation
ou
corporation selects and publishes bench marks.
The program selected range from game playing, compiler, and data base
gr
applications to numerically intensive programs in astrophysics and quantum chemistry. In
each case, the program is compiled under test, and the running time on a real computer is
ts
measured. The same program is also compiled and run on one computer selected as
reference.
en
The SPEC rating is computed as follows.
Running time on the reference computer

ud
SPEC rating =
Running time on the computer under test
st
If the SPEC rating = 50

ity
Means that the computer under test is 50 times as fast as the ultra sparc 10. This
is repeated for all the programs in the SPEC suit, and the geometric mean of the result is
.c
computed.
w
Let SPECi be the rating for program i in the suite. The overall SPEC rating for
the computer is given by
w
( )
n 1
n
SPEC rating = SP ECi
i= 1
w
Where n = number of programs in suite.

Since actual execution time is measured the SPEC rating is a measure of the
combined effect of all factors affecting performance, including the compiler, the OS, the
processor, the memory of comp being tested.
Multiprocessor & microprocessors:-

Large computers that contain a number of processor units are called
om
multiprocessor system.
These systems either execute a number of different application tasks in parallel or
c
execute subtasks of a single large task in parallel.
All processors usually have access to all memory locations in such system &
p.
hence they are called shared memory multiprocessor systems.
The high performance of these systems comes with much increased complexity
ou
and cost.
In contrast to multiprocessor systems, it is also possible to use an interconnected
group of complete computers to achieve high total computational power. These
gr
computers normally have access to their own memory units when the tasks they
are executing need to communicate data they do so by exchanging messages over
ts
a communication network. This properly distinguishes them from shared memory
multiprocessors, leading to name message-passing multi computer.
en
1.10 Number Representation

ud
Consider an n-bit vector

B = bn1

b1 b0
st
Where bi = 0 or 1 for 0 i n1
. This vector can represent unsigned integer values V in
n
the range 0 to 2 -1, where
ity
V(B) = bn1 2 +
n1
+b1 21 +b0 20
We obviously need to represent both positive and negative numbers. Three systems are
.c
used for representing such numbers :

Sign-and-magnitude
w
1s-complement
w
2s-complement
In all three systems, the leftmost bit is 0 for positive numbers and 1 for negative numbers.
w
Fig 2.1 illustrates all three representations using 4-bit numbers. Positive values have
identical representations in al systems, but negative values have different representations.
In the sign-and-magnitude systems, negative values are represented by changing the most
significant bit (b3 in figure 2.1) from 0 to 1 in the B vector of the corresponding positive
value. For example, +5 is represented by 0101, and -5 is represented by 1101. In 1s-

complement representation, negative values are obtained by complementing each bit of

the corresponding positive number. Thus, the representation for -3 is obtained by
complementing each bit in the vector 0011 to yield 1100. clearly, the same operation, bit
complementing, is done in converting a negative number to the corresponding positive
value. Converting either way is referred to as forming the 1s-complement of a given
number. Finally, in the 2s-complement system, forming the 2s-complement of a number
om
is done by subtracting that number from 2n.
B Values represented
c
Sign and
b3b2b1 1's 2's
p.
b0 magnitude complement complement
0 1 1
ou
1 +7 +7 +7
0 1 1
0 +6 +6 +6
gr
0 1 0
1 +5 ts +5 +5
0 1 0
0 +4 +4 +4
en
0 0 1
1 +3 +3 +3
0 0 1
ud
0 +2 +2 +2
0 0 0
1 +1 +1 +1
st
0 0 0
0 +0 +0 +0
ity
1 0 0
0 -0 -0 -0
.c
1 0 0
1 -1 -1 -1
w
1 0 1
0 -2 -2 -2
w
1 0 1
1 -3 -3 -3
w
1 1 0
0 -4 -4 -4
1 1 0
1 -5 -5 -5

1 1 1
0 -6 -6 -6
1 1 1
1 -7 -7 -7
Hence, the 2s complement of a number is obtained by adding 1 to the 1s complement of
that number.
om
Addition of Positive numbers:-
Consider adding two 1-bit numbers. The results are shown in figure 2.2. Note
c
that the sum of 1 and 1 requires the 2-bit vector 10 to represent the value 2. We say that
the sum is 0 and the carry-out is 1. In order to add multiple-bit numbers, we use a method
p.
analogous to that used for manual computation with decimal numbers. We add bit pairs
starting from the low-order (right) and of the bit vectors, propagating carries toward the
ou
high-order (left) end.
0 1 0 1
gr
+0 +0 +1 +1
____ ____ ts ___ ___
0 1 1 10
en
Carry-out
Figure 2.2 Addition of 1-bit numbers.
ud
1.12 Memory locations and addresses

st
Number and character operands, as well as instructions, are stored in the memory
of a computer. The memory consists of many millions of storage cells, each of which can
ity
store a bit of information having the value 0 or 1. Because a single bit represents a very
small amount of information, bits are seldom handled individually. The usual approach is
.c
to deal with them in groups of fixed size. For this purpose, the memory is organized so
that a group of n bits can be stored or retrieved in a single, basic operation. Each group of
w
n bits is referred to as a word of information, and n is called the word length. The
memory of a computer can be schematically represented as a collection of words as
w
shown in figure (a).

w
Modern computers have word lengths that typically range from 16 to 64 bits. If
the word length of a computer is 32 bits, a single word can store a 32-bit 2s complement
number or four ASCII characters, each occupying 8 bits. A unit of 8 bits is called a byte.

Accessing the memory to store or retrieve a single item of information, either a

word or a byte, requires distinct names or addresses for each item location. It is
customary to use numbers from 0 through 2K-1, for some suitable values of k, as the
addresses of successive locations in the memory. The 2k addresses constitute the address
space of the computer, and the memory can have up to 2k addressable locations. 24-bit
address generates an address space of 224 (16,777,216) locations. A 32-bit address creates
om
an address space of 232 or 4G (4 giga) locations.
BYTE ADDRESSABILITY:-
c
We now have three basic information quantities to deal with: the bit, byte and
word. A byte is always 8 bits, but the word length typically ranges from 16 to 64 bits.
p.
The most practical assignment is to have successive addresses refer to successive byte
ou
Fig a Memory words
n bits
gr
First word
ts Second word

en

ud
i-th word

st

ity
Last word
.c
(a) A signed integer

w
w
32 bits
b31 b30 . b1 b0
w
Sign bit: b31 = 0 for positive numbers

b32 = 1 for negative numbers

(b) Four characters
8 bits 8 bits 8 bits 8 bits

ASCII ASCII ASCII ASCII
Character character character character
om
Locations in the memory. This is the assignment used in most modern computers, and is
the one we will normally use in this book. The term byte-addressable memory is use for
this assignment. Byte locations have addresses 0,1,2, . Thus, if the word length of the
machine is 32 bits, successive words are located at addresses 0,4,8,., with each word
c
consisting of four bytes.
p.
BIG-ENDIAN AND LITTLE-ENDIAN ASIGNMENTS:-
ou
There are two ways that byte addresses can be assigned across words, as shown
in fig b. The name big-endian is used when lower byte addresses are used for the more
significant bytes (the leftmost bytes) of the word. The name little-endian is used for the
gr
opposite ordering, where the lower byte addresses are used for the less significant bytes
(the rightmost bytes) of the word. ts
In addition to specifying the address ordering of bytes within a word, it is also
en
necessary to specify the labeling of bits within a byte or a word. The same ordering is
also used for labeling bits within a byte, that is, b7, b6, ., b0, from left to right.
Word
ud
Address Byte address Byte address
0 1 2 3 3 2 1 0
0 0
st
4 5 6 7 7 6 5 4
4 4
ity
. .
.
.c
.
. .
w
w
2k-4 2k-3 2k-2 2k-1 2k-1 2k-2 2k-3 2k-4

w
2k-4 2k-4
(a) Big-endian assignment (b) Little-endian assignment
Dept of Page 22
CSE,SJBIT
WORD ALIGNMENT:-
In the case of a 32-bit word length, natural word boundaries occur at addresses 0,
4, 8, , as shown in above fig. We say that the word locations have aligned addresses .
in general, words are said to be aligned in memory if they begin at a byte address that is a
multiple of the number of bytes in a word. The memory of bytes in a word is a power of
om
2. Hence, if the word length is 16 (2 bytes), aligned words begin at byte addresses
0,2,4,, and for a word length of 64 (23 bytes), aligned words begin at bytes addresses
0,8,16 .
c
There is no fundamental reason why words cannot begin at an arbitrary byte
p.
address. In that case, words are said to have unaligned addresses. While the most
common case is to use aligned addresses, some computers allow the use of unaligned
ou
word addresses.
gr
ACCESSING NUMBERS, CHARACTERS, AND CHARACTER STRINGS:-
A number usually occupies one word. It can be accessed in the memory by
ts
specifying its word address. Similarly, individual characters can be accessed by their byte
address.
en
In many applications, it is necessary to handle character strings of variable
length. The beginning of the string is indicated by giving the address of the byte
ud
containing its first character. Successive byte locations contain successive characters of
the string. There are two ways to indicate the length of the string. A special control
character with the meaning end of string can be used as the last character in the string,
st
or a separate memory word location or processor register can contain a number indicating
the length of the string in bytes.
ity
1.13 Memory operations

.c
Both program instructions and data operands are stored in the memory. To
execute an instruction, the processor control circuits must cause the word (or words)
w
containing the instruction to be transferred from the memory to the processor. Operands
and results must also be moved between the memory and the processor. Thus, two basic
w
operations involving the memory are needed, namely, Load (or Read or Fetch) and Store
(or Write).
w
The load operation transfers a copy of the contents of a specific memory location
to the processor. The memory contents remain unchanged. To start a Load operation, the
processor sends the address of the desired location to the memory and requests that its

contents be read. The memory reads the data stored at that address and sends them to the
processor.
The store operation transfers an item of information from the processor to a

specific memory location, destroying the former contents of that location. The processor
sends the address of the desired location to the memory, together with the data to be
om
written into that location.
An information item of either one word or one byte can be transferred between
c
the processor and the memory in a single operation. Actually this transfer in between the
CPU register & main memory.
p.
ou
1.14 Instructions and instruction sequencing
gr
A computer must have instructions capable of performing four types of
operations.
Data transfers between the memory and the processor registers
ts
Arithmetic and logic operations on data
Program sequencing and control
en
I/O transfers
ud
REGISTER TRANSFER NOTATION:-

Transfer of information from one location in the computer to another. Possible
locations that may be involved in such transfers are memory locations that may be
st
involved in such transfers are memory locations, processor registers, or registers in the
I/O subsystem. Most of the time, we identify a location by a symbolic name standing for
ity
its hardware binary address.
Example, names for the addresses of memory locations may be LOC, PLACE, A,
.c
VAR2; processor registers names may be R0, R5; and I/O register names may be
DATAIN, OUTSTATUS, and so on. The contents of a location are denoted by placing
w
square brackets around the name of the location. Thus, the expression
R1 [LOC]
w
Means that the contents of memory location LOC are transferred into processor register
w
R1.
As another example, consider the operation that adds the contents of registers R1
and R2, and then places their sum into register R3. This action is indicated as
R3 [R1] + [R2]

This type of notation is known as Register Transfer Notation (RTN). Note that
the right-hand side of an RTN expression always denotes a value, and the left-hand side
is the name of a location where the value is to be places, overwriting the old contents of
that location.
om
ASSEMBLY LANGUAGE NOTATION:-
Another type of notation to represent machine instructions and programs. For
this, we use an assembly language format. For example, an instruction that causes the
c
transfer described above, from memory location LOC to processor register R1, is
specified by the statement
p.
Move LOC, R1
ou
The contents of LOC are unchanged by the execution of this instruction, but the
old contents of register R1 are overwritten.
gr
The second example of adding two numbers contained in processor registers R1
and R2 and placing their sum in R3 can be specified by the assembly language statement
ts
Add R1, R2, R3
en
BASIC INSTRUCTIONS:-
The operation of adding two numbers is a fundamental capability in any
computer. The statement
ud
C=A+B
In a high-level language program is a command to the computer to add the

st
current values of the two variables called A and B, and to assign the sum to a third
variable, C. When the program containing this statement is compiled, the three variables,
ity
A, B, and C, are assigned to distinct locations in the memory. We will use the variable
names to refer to the corresponding memory location addresses. The contents of these
.c
locations represent the values of the three variables. Hence, the above high-level
language statement requires the action.
w
C [A] + [B]
w
To carry out this action, the contents of memory locations A and B are fetched
from the memory and transferred into the processor where their sum is computed. This
w
result is then sent back to the memory and stored in location C.
Let us first assume that this action is to be accomplished by a single machine

instruction. Furthermore, assume that this instruction contains the memory addresses of

the three operands A, B, and C. This three-address instruction can be represented

symbolically as
Add A, B, C
Operands A and B are called the source operands, C is called the destination
operand, and Add is the operation to be performed on the operands. A general instruction
om
of this type has the format.
Operation Source1, Source 2, Destination
c
If k bits are needed for specify the memory address of each operand, the encoded
form of the above instruction must contain 3k bits for addressing purposes in addition to
p.
the bits needed to denote the Add operation.
ou
An alternative approach is to use a sequence of simpler instructions to perform
the same task, with each instruction having only one or two operands. Suppose that two-
address instructions of the form
gr
Operation Source, Destination
ts
Are available. An Add instruction of this type is
Add A, B
en
Which performs the operation B [A] + [B].
ud
A single two-address instruction cannot be used to solve our original problem,

which is to add the contents of locations A and B, without destroying either of them, and
to place the sum in location C. The problem can be solved by using another two-address
st
instruction that copies the contents of one memory location into another. Such an
instruction is
ity
Move B, C
.c

Which performs the operations C [B], leaving the contents of location B unchanged.
w

Using only one-address instructions, the operation C [A] + [B] can be
performed by executing the sequence of instructions
w
Load A
Add B
w
Store C
Some early computers were designed around a single accumulator structure.

Most modern computers have a number of general-purpose processor registers typically

8 to 32, and even considerably more in some cases. Access to data in these registers is
much faster than to data stored in memory locations because the registers are inside the
processor.
Let Ri represent a general-purpose register. The instructions

Load A, Ri
om
Store Ri, A and
Add A, Ri
c
Are generalizations of the Load, Store, and Add instructions for the single-accumulator
case, in which register Ri performs the function of the accumulator.
p.
When a processor has several general-purpose registers, many instructions
ou
involve only operands that are in the register. In fact, in many modern processors,
computations can be performed directly only on data held in processor registers.
Instructions such as
gr
Add Ri, Rj
ts Or
Add Ri, Rj, Rk
In both of these instructions, the source operands are the contents of registers Ri
en
and Rj. In the first instruction, Rj also serves as the destination register, whereas in the
second instruction, a third register, Rk, is used as the destination.
ud
It is often necessary to transfer data between different locations. This is achieved

with the instruction
Move Source, Destination
st
When data are moved to or from a processor register, the Move instruction can be
used rather than the Load or Store instructions because the order of the source and
ity
destination operands determines which operation is intended. Thus,

Move A, Ri
.c
Is the same as
Load A, Ri
w
And
Move Ri, A
w
Is the same as
Store Ri, A
w
In processors where arithmetic operations are allowed only on operands that are
processor registers, the C = A + B task can be performed by the instruction sequence
Move A, Ri
Move B, Rj

Add Ri, Rj
Move Rj, C
In processors where one operand may be in the memory but the other must be in
register, an instruction sequence for the required task would be
Move A, Ri
Add B, Ri
om
Move Ri, C
The speed with which a given task is carried out depends on the time it takes to
transfer instructions from memory into the processor and to access the operands
c
referenced by these instructions. Transfers that involve the memory are much slower than
transfers within the processor.
p.
We have discussed three-, two-, and one-address instructions. It is also possible
to use instructions in which the locations of all operands are defined implicitly. Such
ou
instructions are found in machines that store operands in a structure called a pushdown
stack. In this case, the instructions are called zero-address instructions.
gr
INSTRUCTION EXECUTION AND STRAIGHT-LINE SEQUENCING:-
In the preceding discussion of instruction formats, we used to task C [A] +
ts
[B]. fig 2.8 shows a possible program segment for this task as it appears in the memory of
a computer. We have assumed that the computer allows one memory operand per
en
instruction and has a number of processor registers. The three instructions of the program
are in successive word locations, starting at location i. since each instruction is 4 bytes
long, the second and third instructions start at addresses i + 4 and i + 8.
ud
Address Contents
Begin execution here i 3-instruction

st
Move A,
i+4 R0 program
Move B, segment
ity
i+8 R0Move R0, C

.c

w
w
A

w
B Data for the

program

Let us consider how this program is executed. The processor contains a register
called the program counter (PC), which holds the address of the instruction to be
executed next. To begin executing a program, the address of its first instruction (I in our
example) must be placed into the PC. Then, the processor control circuits use the
information in the PC to fetch and execute instructions, one at a time, in the order of
increasing addresses. This is called straight-line sequencing. During the execution of each
om
instruction, the PC is incremented by 4 to point to the next instruction. Thus, after the
Move instruction at location i + 8 is executed, the PC contains the value i + 12, which is
the address of the first instruction of the next program segment.
c
Executing a given instruction is a two-phase procedure. In the first phase, called
p.
instruction fetch, the instruction is fetched from the memory location whose address is in
the PC. This instruction is placed in the instruction register (IR) in the processor. The
ou
instruction in IR is examined to determine which operation is to be performed. The
specified operation is then performed by the processor. This often involves fetching
operands from the memory or from processor registers, performing an arithmetic or logic
gr
operation, and storing the result in the destination location.
BRANCHING:-
ts
Consider the task of adding a list of n numbers. Instead of using a long list of add
en
instructions, it is possible to place a single add instruction in a program loop, as shown in
fig b. The loop is a straight-line sequence of instructions executed as many times as
needed. It starts at location LOOP and ends at the instruction Branch > 0. During each
ud
pass through this loop, the address of the next list entry is determined, and that entry is
fetched and added to
st
Move NUM1, R0 i
Add NUM2, R0 i+4
ity
Add NUM3, R0 i+8

.c
Add NUMn, R0
Move R0, SUM
w
i+4n-4
. i+4n
w
Dept of . Page 30
CSE,SJBIT
fig a A straight-line program for adding n numbers
Move N, R1
Clear R0
om
Determine address of
Next number and add LOOP
Next number to R0
c
Decrement R1 Program
Branch >0 LOOP loop
p.
Move R0, SUM
ou
.
.
.
gr
ts
en

...
ud
SUM N
NUM1 NUM2
st
NUMn
Fig b Using a loop to add n numbers
ity
Assume that the number of entries in the list, n, is stored in memory location N,
.c
as shown. Register R1 is used as a counter to determine the number of time the loop is
executed. Hence, the contents of location N are loaded into register R1 at the beginning
w
of the program. Then, within the body of the loop, the instruction.
Decrement R1
w
Reduces the contents of R1 by 1 each time through the loop.

This type of instruction loads a new value into the program counter. As a result,
w
the processor fetches and executes the instruction at this new address, called the branch
target, instead of the instruction at the location that follows the branch instruction in
sequential address order. A conditional branch instruction causes a branch only if a
specified condition is satisfied. If the condition is not satisfied, the PC is incremented in

the normal way, and the next instruction in sequential address order is fetched and
executed.
Branch > 0 LOOP
(branch if greater that 0) is a conditional branch instruction that causes a branch

to location LOOP if the result of the immediately preceding instruction, which is the
om
decremented value in register R1, is greater that zero. This means that the loop is
repeated, as long as there are entries in the list that are yet to be added to R0. at the end of
the nth pass through the loop, the Decrement instruction produces a value of zero, and
c
hence, branching does not occur.
p.
CONDITION CODES:-
ou
The processor keeps track of information about the results of various operations
for use by subsequent conditional branch instructions. This is accomplished by recording
the required information in individual bits, often called condition code flags. These flags
gr
are usually grouped together in a special processor register called the condition code
register or status register. Individual condition code flags are set to 1 or cleared to 0,
ts
depending on the outcome of the operation performed.
en
Four commonly used flags are
N(negative) Set to 1 if the result is negative; otherwise, cleared to 0

ud
Z(zero) Set to 1 if the result is 0; otherwise, cleared to 0

V(overflow) Set ot1 if arithmetic overflow occurs; otherwise, cleared to 0
C(carry) Set to 1 if a carry-out results from the operation; otherwise, cleared to 0
st
The instruction Branch > 0, discussed in the previous section, is an example of a

ity
branch instruction that tests one or more of the condition flags. It causes a branch if the
value tested is neither negative nor equal to zero. That is, the branch is taken if neither N
.c
nor Z is 1. The conditions are given as logic expressions involving the condition code
flags.
w
In some computers, the condition code flags are affected automatically by

w
instructions that perform arithmetic or logic operations. However, this is not always the
case. A number of computers have two versions of an Add instruction.
w
GENERATING MEMORY ADDRESSES:-

Let us return to fig b. The purpose of the instruction block at LOOP is to add a
different number from the list during each pass through the loop. Hence, the Add
instruction in the block must refer to a different address during each pass. How are the

addresses to be specified ? The memory operand address cannot be given directly in a

single Add instruction in the loop. Otherwise, it would need to be modified on each pass
through the loop.
The instruction set of a computer typically provides a number of such methods,

called addressing modes. While the details differ from one computer to another, the
om
underlying concepts are the same.
c
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
w

UNIT - 2
Machine Instructions and Programs contd.: Addressing Modes,
Assembly Language, Basic Input and Output Operations, Stacks and
Queues, Subroutines, Additional Instructions, Encoding of Machine
om
Instructions
c
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
w

CHAPTER 2
MACHINE INSTRUCTIONS AND PROGRAMS CONTD.:
2.1 Addressing modes:
om
In general, a program operates on data that reside in the computers memory.
These data can be organized in a variety of ways. If we want to keep track of students
names, we can write them in a list. Programmers use organizations called data structures
c
to represent the data used in computations. These include lists, linked lists, arrays,
p.
queues, and so on.
Programs are normally written in a high-level language, which enables the
ou
programmer to use constants, local and global variables, pointers, and arrays. The
different ways in which the location of an operand is specified in an instruction are
gr
referred to as addressing modes.
Table 2.1 Generic addressing modes

ts
Name Assembler syntax Addressing function
en
Immediate # Value Operand = Value
ud
Register Ri EA = Ri
Absolute (Direct) LOC EA = LOC
Indirect (Ri) EA = [Ri]
st
(LOC) EA = [LOC]
Index X(Ri) EA = [Ri] + X
ity
Base with index (Ri, Rj) EA = [Ri] + [Rj]

Base with index X (Ri, Rj) EA = [Ri] + [Rj] + X
and offset
.c
Relative X(PC) EA = [PC] + X

Autoincrement (Ri)+ EA = [Ri]; Increment Ri
w
Autodecrement -(Ri) Decrement Ri; EA = [Ri]

w
EA = effective address
Value = a signed number
w

IMPLEMENTATION OF VARIABLE AND CONSTANTS:-

Variables and constants are the simplest data types and are found in almost every
computer program. In assembly language, a variable is represented by allocating a
register or memory location to hold its value. Thus, the value can be changed as needed
using appropriate instructions.
om
Register mode - The operand is the contents of a processor register; the name (address)
of the register is given in the instruction.
Absolute mode The operand is in a memory location; the address of this location is
c
given explicitly in the instruction. (In some assembly languages, this mode is called
p.
Direct).
ou
The instruction
Move LOC, R2
gr
Processor registers are used as temporary storage locations where the data is a
register are accessed using the Register mode. The Absolute mode can represent global
ts
variables in a program. A declaration such as
Integer A, B;
en
Immediate mode The operand is given explicitly in the instruction.
For example, the instruction
ud
Move 200immediate, R0
Places the value 200 in register R0. Clearly, the Immediate mode is only used to
st
specify the value of a source operand. Using a subscript to denote the Immediate mode is
not appropriate in assembly languages. A common convention is to use the sharp sign (#)
ity
in front of the value to indicate that this value is to be used as an immediate operand.
Hence, we write the instruction above in the form
.c
Move #200, R0
w
INDIRECTION AND POINTERS:-

In the addressing modes that follow, the instruction does not give the operand or
w
its address explicitly, Instead, it provides information from which the memory address of
the operand can be determined. We refer to this address as the effective address (EA) of
w
the operand.
Indirect mode The effective address of the operand is the contents of a register or
memory location whose address appears in the instruction.

To execute the Add instruction in fig (a), the processor uses the value which is in
register R1, as the effective address of the operand. It requests a read operation from the
memory to read the contents of location B. the value read is the desired operand, which
the processor adds to the contents of register R0. Indirect addressing through a memory
location is also possible as shown in fig (b). In this case, the processor first reads the
contents of memory location A, then requests a second read operation using the value B
om
as an address to obtain the operand
Fig (a) Through a general-purpose register (b) Through a memory location
c
Add (A), R0
p.
Add (R1), R0

ou
Main

memory
B
A
gr
Operand
ts
B
en
Operands
R1 B Register
B
ud
Address Contents
st
Move N, R1
ity
Move #NUM, R2
Clear R0
.c
LOOP Add (R2), R0

Add #4, R2
w
Decrement R1
Branch > 0 LOOP
w
Move R0, SUM

w
The register or memory location that contains the address of an operand is called
a pointer. Indirection and the use of pointers are important and powerful concepts in
programming.

In the program shown Register R2 is used as a pointer to the numbers in the list,
and the operands are accessed indirectly through R2. The initialization section of the
program loads the counter value n from memory location N into R1 and uses the
immediate addressing mode to place the address value NUM1, which is the address of the
first number in the list, into R2. Then it clears R0 to 0. The first two instructions in the
loop implement the unspecified instruction block starting at LOOP. The first time
om
through the loop, the instruction Add (R2), R0 fetches the operand at location NUM1
and adds it to R0. The second Add instruction adds 4 to the contents of the pointer R2, so
that it will contain the address value NUM2 when the above instruction is executed in the
c
second pass through the loop.
p.
Where B is a pointer variable. This statement may be compiled into
Move B, R1
ou
Move (R1), A
Using indirect addressing through memory, the same action can be achieved with
Move (B), A
gr
Indirect addressing through registers is used extensively. The above program
ts
shows the flexibility it provides. Also, when absolute addressing is not available, indirect
addressing through registers makes it possible to access global variables by first loading
en
the operands address in a register.
INDEXING AND ARRAYS:-

ud
A different kind of flexibility for accessing operands is useful in dealing with

lists and arrays.
st
Index mode the effective address of the operand is generated by adding a constant
value to the contents of a register.
ity
The register use may be either a special register provided for this purpose, or,
.c
more commonly, it may be any one of a set of general-purpose registers in the processor.
In either case, it is referred to as index register. We indicate the Index mode symbolically
w
as
X (Ri)
w
Where X denotes the constant value contained in the instruction and Ri is the
w
name of the register involved. The effective address of the operand is given by
EA = X + [Rj]

The contents of the index register are not changed in the process of generating
the effective address. In an assembly language program, the constant X may be given
either as an explicit number or as a symbolic name representing a numerical value.
Fig a illustrates two ways of using the Index mode. In fig a, the index register,
R1, contains the address of a memory location, and the value X defines an offset (also
om
called a displacement) from this address to the location where the operand is found. An
alternative use is illustrated in fig b. Here, the constant X corresponds to a memory
address, and the contents of the index register define the offset to the operand. In either
c
case, the effective address is the sum of two values; one is given explicitly in the
instruction, and the other is stored in a register.
p.
Fig (a) Offset is given as a constant
ou
Add 20(R1), R2

gr

ts
1000 1000 R1

en

20 = offset
ud
Operands
1020
st
Add 1000(R1), R2
ity

.c
20
w
1000 R1

w

20 = offset
w
Operand
1020
Fig (b) Offset is in the index register

Move #LIST, R0
Clear R1
Clear R2
Clear R3
Move N, R4
om
LOOP Add 4(R0), R1
Add 8(R0), R2
Add 12(R0), R3
c
Add #16, R0
Decrement R4
p.
Branch>0 LOOP
Move R1, SUM1
ou
Move R2, SUM2
Move R3, SUM3
gr
In the most basic form of indexed addressing several variations of this basic
ts
form provide a very efficient access to memory operands in practical programming
situations. For example, a second register may be used to contain the offset X, in which
en
case we can write the Index mode as
(Ri, Rj)
ud
The effective address is the sum of the contents of registers Ri and Rj. The
second register is usually called the base register. This form of indexed addressing
st
provides more flexibility in accessing operands, because both components of the effective
address can be changed.
ity
Another version of the Index mode uses two registers plus a constant, which can
.c
be denoted as
w
X(Ri, Rj)
w
In this case, the effective address is the sum of the constant X and the contents of
registers Ri and Rj. This added flexibility is useful in accessing multiple components
w
inside each item in a record, where the beginning of an item is specified by the (Ri, Rj)
part of the addressing mode. In other words, this mode implements a three-dimensional
array.

RELATIVE ADDRESSING:-
We have defined the Index mode using general-purpose processor registers. A
useful version of this mode is obtained if the program counter, PC, is used instead of a
general purpose register. Then, X(PC) can be used to address a memory location that is X
bytes away from the location presently pointed to by the program counter.
om
Relative mode The effective address is determined by the Index mode using the
program counter in place of the general-purpose register Ri.
c
This mode can be used to access data operands. But, its most common use is to
specify the target address in branch instructions. An instruction such as
p.
Branch > 0 LOOP
ou
Causes program execution to go to the branch target location identified by the
name LOOP if the branch condition is satisfied. This location can be computed by
gr
specifying it as an offset from the current value of the program counter. Since the branch
target may be either before or after the branch instruction, the offset is given as a signed
ts
number.
en
Autoincrement mode the effective address of the operand is the contents of a register
specified in the instruction. After accessing the operand, the contents of this register are
automatically to point to the next item in a list.
ud
(Ri)+
st
Autodecrement mode the contents of a register specified in the instruction are first
automatically decremented and are then used as the effective address of the operand.
ity
-(Ri)
.c
Move N, R1
Move #NUM1, R2
w
Clear R0
LOOP Add (R2)+, R0
w
Decrement R1
Branch>0 LOOP
w
Move R0, SUM
Fig c The Autoincrement addressing mode used in the program of fig 2.12

2.2 ASSEMBLY LANGUAGE

Machine instructions are represented by patterns of 0s and 1s. Such patterns are
awkward to deal with when discussing or preparing programs. Therefore, we use
symbolic names to represent the pattern. So far, we have used normal words, such as
Move, Add, Increment, and Branch, for the instruction operations to represent the
corresponding binary code patterns. When writing programs for a specific computer, such
om
words are normally replaced by acronyms called mnemonics, such as MOV, ADD, INC,
and BR. Similarly, we use the notation R3 to refer to register 3, and LOC to refer to a
memory location. A complete set of such symbolic names and rules for their use
c
constitute a programming language, generally referred to as an assembly language.
p.
Programs written in an assembly language can be automatically translated into a
sequence of machine instructions by a program called an assembler. When the assembler
ou
program is executed, it reads the user program, analyzes it, and then generates the desired
machine language program. The latter contains patterns of 0s and 1s specifying
instructions that will be executed by the computer. The user program in its original
gr
alphanumeric text format is called a source program, and the assembled machine
language program is called an object program.
ts
ASSEMBLER DIRECTIVES:-
en
In addition to providing a mechanism for representing instructions in a program,
the assembly language allows the programmer to specify other information needed to
translate the source program into the object program. We have already mentioned that we
ud
need to assign numerical values to any names used in a program. Suppose that the name
SUM is used to represent the value 200. This fact may be conveyed to the assembler
program through a statement such as
st
SUM EQU 200

ity
This statement does not denote an instruction that will be executed when the
.c
object program is run; in fact, it will not even appear in the object program. It simply
informs the assembler that the name SUM should be replaced by the value 200 wherever
w
it appears in the program. Such statements, called assembler directives (or commands),
are used by the assembler while it translates a source program into an object program.
w
w

Move N, R1 100
Move # NUM1,R2 104
Clear R0
Add (R2), R0 108
Add #4, R2 LOOP 112
Decrement R1 116
om
Branch>0 LOOP
Move R0, SUM 120
124
c
. 128
.
p.
. 132
ou
100
gr
SUM 200
.
N
.
ts 204
NUM1 208 NUM2 212
en
NUMn 604
ud
Fig 2.17 Memory arrangement for the program in fig b.

st
ASSEMBLY AND EXECUTION OF PRGRAMS:-

A source program written in an assembly language must be assembled into a
ity
machine language object program before it can be executed. This is done by the
assembler program, which replaces all symbols denoting operations and addressing
.c
modes with the binary codes used in machine instructions, and replaces all names and
labels with their actual values.
w
The assembler assigns addresses to instructions and data blocks, starting at the
w
address given in the ORIGIN assembler directives. It also inserts constants that may be
given in DATAWORD commands and reserves memory space as requested by
w
RESERVE commands.
As the assembler scans through a source programs, it keeps track of all names
and the numerical values that correspond to them in a symbol table. Thus, when a name
appears a second time, it is replaced with its value from the table. A problem arises when

a name appears as an operand before it is given a value. For example, this happens if a
forward branch is required. A simple solution to this problem is to have the assembler
scan through the source program twice. During the first pass, it creates a complete
symbol table. At the end of this pass, all names will have been assigned numerical values.
The assembler then goes through the source program a second time and substitutes values
for all names from the symbol table. Such an assembler is called a two-pass assembler.
om
The assembler stores the object program on a magnetic disk. The object program
must be loaded into the memory of the computer before it is executed. For this to happen,
c
another utility program called a loader must already be in the memory.
p.
When the object program begins executing, it proceeds to completion unless
there are logical errors in the program. The user must be able to find errors easily. The
ou
assembler can detect and report syntax errors. To help the user find other programming
errors, the system software usually includes a debugger program. This program enables
the user to stop execution of the object program at some points of interest and to examine
gr
the contents of various processor registers and memory locations.
NUMBER NOTATION:-
ts
When dealing with numerical values, it is often convenient to use the familiar
en
decimal notation. Of course, these values are stored in the computer as binary numbers.
In some situations, it is more convenient to specify the binary patterns directly. Most
assemblers allow numerical values to be specified in different ways, using conventions
ud
that are defined by the assembly language syntax. Consider, for example, the number 93,
which is represented by the 8-bit binary number 01011101. If this value is to be used an
immediate operand, it can be given as a decimal number, as in the instructions.
st
ADD #93, R1
ity
Or as a binary number identified by a prefix symbol such as a percent sign, as in

.c
ADD #%01011101, R1
w
Binary numbers can be written more compactly as hexadecimal, or hex, numbers,

w
in which four bits are represented by a single hex digit. In hexadecimal representation,
the decimal value 93 becomes 5D. In assembly language, a hex representation is often
w
identified by a dollar sign prefix. Thus, we would write
ADD #$5D, R1

2.3 Basic input/output operations

We now examine the means by which data are transferred between the memory
of a computer and the outside world. Input/Output (I/O) operations are essential, and the
way they are performed can have a significant effect on the performance of the computer.
Consider a task that reads in character input from a keyboard and produces
om
character output on a display screen. A simple way of performing such I/O tasks is to use
a method known as program-controlled I/O. The rate of data transfer from the keyboard
to a computer is limited by the typing speed of the user, which is unlikely to exceed a few
c
characters per second. The rate of output transfers from the computer to the display is
much higher. It is determined by the rate at which characters can be transmitted over the
p.
link between the computer and the display device, typically several thousand characters
per second. However, this is still much slower than the speed of a processor that can
ou
execute many millions of instructions per second. The difference in speed between the
processor and I/O devices creates the need for mechanisms to synchronize the transfer of
data between them.
gr
ts Bus
en
ud
Processor DATAIN DATAOUT

st
SIN SOUT
ity
Keyboard Display
.c
Fig a Bus connection for processor, keyboard, and display

w
The keyboard and the display are separate device as shown in fig a. the action of
w
striking a key on the keyboard does not automatically cause the corresponding character
to be displayed on the screen. One block of instructions in the I/O program transfers the
w
character into the processor, and another associated block of instructions causes the
character to be displayed.
Striking a key stores the corresponding character code in an 8-bit buffer register
associated with the keyboard. Let us call this register DATAIN, as shown in fig a. To

inform the processor that a valid character is in DATAIN, a status control flag, SIN, is set
to 1. A program monitors SIN, and when SIN is set to 1, the processor reads the contents
of DATAIN. When the character is transferred to the processor, SIN is automatically
cleared to 0. If a second character is entered at the keyboard, SIN is again set to 1, and the
processor repeats.
om
An analogous process takes place when characters are transferred from the
processor to the display. A buffer register, DATAOUT, and a status control flag, SOUT,
are used for this transfer. When SOUT equals 1, the display is ready to receive a
c
character.
p.
In order to perform I/O transfers, we need machine instructions that can check
the state of the status flags and transfer data between the processor and the I/O device.
ou
These instructions are similar in format to those used for moving data between the
processor and the memory. For example, the processor can monitor the keyboard status
flag SIN and transfer a character from DATAIN to register R1 by the following sequence
gr
of operations.
2.4 Stacks and queues

ts
A computer program often needs to perform a particular subtask using the
en
familiar subroutine structure. In order to organize the control and information linkage
between the main program and the subroutine, a data structure called a stack is used. This
ud
section will describe stacks, as well as a closely related data structure called a queue.
st
Data operated on by a program can be organized in a variety of ways. We have

ity
already encountered data structured as lists. Now, we consider an important data structure
known as a stack. A stack is a list of data elements, usually words or bytes, with the
.c
accessing restriction that elements can be added or removed at one end of the list only.
This end is called the top of the stack, and the other end is called the bottom. Another
w
descriptive phrase, last-in-first-out (LIFO) stack, is also used to describe this type of
w
storage mechanism; the last data item placed on the stack is the first one removed when
retrieval begins. The terms push and pop are used to describe placing a new item on the
w
stack and removing the top item from the stack, respectively.

Fig b shows a stack of word data items in the memory of a computer. It contains
numerical values, with 43 at the bottom and -28 at the top. A processor register is used to
keep track of the address of the element of the stack that is at the top at any given time.
This register is called the stack pointer (SP). It could be one of the general-purpose
registers or a register dedicated to this function.
om
Fig b A stack of words in the memory
c
0
.
p.
. Stack pointer
.
ou
-28 register
17 SP
Current
gr
739
.
.
ts
. Stack
43
en
.
.
.
ud
BOTTOM Bottom element

st
2k-1
ity
Another useful data structure that is similar to the stack is called a queue. Data
are stored in and retrieved from a queue on a first-in-first-out (FIFO) basis. Thus, if we
.c
assume that the queue grows in the direction of increasing addresses in the memory,
which is a common practice, new data are added at the back (high-address end) and
w
retrieved from the front (low-address end) of the queue.

w
There are two important differences between how a stack and a queue are
implemented. One end of the stack is fixed (the bottom), while the other end rises and
w
falls as data are pushed and popped. A single pointer is needed to point to the top of the
stack at any given time. On the other hand, both ends of a queue move to higher
addresses as data are added at the back and removed from the front. So two pointers are
needed to keep track of the two ends of the queue.

Another difference between a stack and a queue is that, without further control, a
queue would continuously move through the memory of a computer in the direction of
higher addresses. One way to limit the queue to a fixed region in memory is to use a
circular buffer. Let us assume that memory addresses from BEGINNING to END are
assigned to the queue. The first entry in the queue is entered into location BEGINNING,
and successive entries are appended to the queue by entering them at successively higher
om
addresses. By the time the back of the queue reaches END, space will have been created
at the beginning if some items have been removed from the queue. Hence, the back
pointer is reset to the value BEGINNING and the process continues. As in the case of a
c
stack, care must be taken to detect when the region assigned to the data structure is either
completely full or completely empty.
p.
2.5 Subroutines
ou
In a given program, it is often necessary to perform a particular subtask many
times on different data-values. Such a subtask is usually called a subroutine. For example,
a subroutine may evaluate the sine function or sort a list of values into increasing or
gr
decreasing order.
ts
It is possible to include the block of instructions that constitute a subroutine at
every place where it is needed in the program. However, to save space, only one copy of
en
the instructions that constitute the subroutine is placed in the memory, and any program
that requires the use of the subroutine simply branches to its starting location. When a
program branches to a subroutine we say that it is calling the subroutine. The instruction
ud
that performs this branch operation is named a Call instruction.
After a subroutine has been executed, the calling program must resume
st
execution, continuing immediately after the instruction that called the subroutine. The
subroutine is said to return to the program that called it by executing a Return instruction.
ity
The way in which a computer makes it possible to call and return from
.c
subroutines is referred to as its subroutine linkage method. The simplest subroutine

linkage method is to save the return address in a specific location, which may be a
w
register dedicated to this function. Such a register is called the link register. When the
subroutine completes its task, the Return instruction returns to the calling program by
w
branching indirectly through the link register.

w
The Call instruction is just a special branch instruction that performs the
following operations
Store the contents of the PC in the link register

Branch to the target address specified by the instruction

The Return instruction is a special branch instruction that performs the operation
Branch to the address contained in the link register .
Fig a illustrates this procedure
om
Memory Memory
Location Calling program location Subroutine SUB
.
.
c
200 Call SUB 1000 first instruction
p.
204 next instruction .
. .
ou
. Return
.
gr
1000
ts
en
204
PC
ud
204
Link
st
Call Return
ity
Fig b Subroutine linkage using a link register

.c
SUBROUTINE NESTING AND THE PROCESSOR STACK:-

w
A common programming practice, called subroutine nesting, is to have one

subroutine call another. In this case, the return address of the second call is also stored in
w
the link register, destroying its previous contents. Hence, it is essential to save the
contents of the link register in some other location before calling another subroutine.
w
Otherwise, the return address of the first subroutine will be lost.
Subroutine nesting can be carried out to any depth. Eventually, the last
subroutine called completes its computations and returns to the subroutine that called it.
The return address needed for this first return is the last one generated in the nested call

sequence. That is, return addresses are generated and used in a last-in-first-out order. This
suggests that the return addresses associated with subroutine calls should be pushed onto
a stack. A particular register is designated as the stack pointer, SP, to be used in this
operation. The stack pointer points to a stack called the processor stack. The Call
instruction pushes the contents of the PC onto the processor stack and loads the
subroutine address into the PC. The Return instruction pops the return address from the
om
processor stack into the PC.
PARAMETER PASSING:-
c
When calling a subroutine, a program must provide to the subroutine the
parameters, that is, the operands or their addresses, to be used in the computation. Later,
p.
the subroutine returns other parameters, in this case, the results of the computation. This
exchange of information between a calling program and a subroutine is referred to as
ou
parameter passing. Parameter passing may be accomplished in several ways. The
parameters may be placed in registers or in memory locations, where they can be
accessed by the subroutine. Alternatively, the parameters may be placed on the processor
gr
stack used for saving the return address.
ts
The purpose of the subroutines is to add a list of numbers. Instead of passing the
actual list entries, the calling program passes the address of the first number in the list.
en
This technique is called passing by reference. The second parameter is passed by value,
that is, the actual number of entries, n, is passed to the subroutine.
ud
THE STACK FRAME:-

Now, observe how space is used in the stack in the example. During execution of
the subroutine, six locations at the top of the stack contain entries that are needed by the
st
subroutine. These locations constitute a private workspace for the subroutine, created at
the time the subroutine is entered and freed up when the subroutine returns control to the
ity
calling program. Such space is called a stack frame.

Fig a A subroutine stack frame example.
.c
Saved [R1] SP
Saved [R0] (stack pointer)
w
Localvar3
Localvar2 Stack
w
Localvar1 frame
Saved [FP]
w
Return address for

Param1 FP called
Param2 (frame pointer) subroutine
P 3

Old TOS
fig b shows an example of a commonly used layout for information in a stack

frame. In addition to the stack pointer SP, it is useful to have another pointer register,
called the frame pointer (FP), for convenient access to the parameters passed to the
om
subroutine and to the local memory variables used by the subroutine. These local
variables are only used within the subroutine, so it is appropriate to allocate space for
them in the stack frame associated with the subroutine. We assume that four parameters
c
are passed to the subroutine, three local variables are used within the subroutine, and
registers R0 and R1 need to be saved because they will also be used within the
p.
subroutine.
ou
The pointers SP and FP are manipulated as the stack frame is built, used, and
dismantled for a particular of the subroutine. We begin by assuming that SP point to the
old top-of-stack (TOS) element in fig b. Before the subroutine is called, the calling
gr
program pushes the four parameters onto the stack. The call instruction is then executed,
resulting in the return address being pushed onto the stack. Now, SP points to this return
ts
address, and the first instruction of the subroutine is about to be executed. This is the
point at which the frame pointer FP is set to contain the proper memory address. Since FP
en
is usually a general-purpose register, it may contain information of use to the Calling
program. Therefore, its contents are saved by pushing them onto the stack. Since the SP
now points to this position, its contents are copied into FP.
ud
Thus, the first two instructions executed in the subroutine are

st
Move FP, -(SP)

Move SP, FP
ity
After these instructions are executed, both SP and FP point to the saved FP contents.
.c
Subtract #12, SP
w
Finally, the contents of processor registers R0 and R1 are saved by pushing them
w
onto the stack. At this point, the stack frame has been set up as shown in the fig.
w
The subroutine now executes its task. When the task is completed, the subroutine
pops the saved values of R1 and R0 back into those registers, removes the local variables
from the stack frame by executing the instruction.
Add #12, SP

And pops the saved old value of FP back into FP. At this point, SP points to the
return address, so the Return instruction can be executed, transferring control back to the
calling program.
2.6 Logic instructions
om
Logic operations such as AND, OR, and NOT, applied to individual bits, are the basic
building blocks of digital circuits, as described. It is also useful to be able to perform
logic operations is software, which is done using instructions that apply these operations
c
to all bits of a word or byte independently and in parallel. For example, the instruction
p.
Not dst
ou
SHIFT AND ROTATE INSTRUCTIONS:-
There are many applications that require the bits of an operand to be shifted right
or left some specified number of bit positions. The details of how the shifts are performed
gr
depend on whether the operand is a signed number or some more general binary-coded
information. For general operands, we use a logical shift. For a number, we use an
ts
arithmetic shift, which preserves the sign of the number.
en
Logical shifts:-
Two logical shift instructions are needed, one for shifting left (LShiftL) and
another for shifting right (LShiftR). These instructions shift an operand over a number of
ud
bit positions specified in a count operand contained in the instruction. The general form
of a logical left shift instruction is
st
LShiftL count, dst

ity
(a) Logical shift left LShiftL #2, R0

.c
R0 0
w
w
0 0 1 1 1 0 . . . 0 1 1
before :
w
after: 1 1 1 0 . . . 0 1 1 00

(b) Logical shift right LShiftR #2, R0
R0 C
om
Before: 0 1 1 1 0 . . . 0 1 1 0
c
p.
1
0 0 0 1 1 1 0 . . . 0
ou
After:
gr
( c) Arithmetic shift right
ts
AShiftR #2, R0
en
R0 C
ud
st
Before: 1 0 0 1 1 . . . 0 1 0 0
ity
.c
1 1 1 0 0 1 1 . . . 0 1
w
After:
w
Rotate Operations:-
w
In the shift operations, the bits shifted out of the operand are lost, except for the
last bit shifted out which is retained in the Carry flag C. To preserve all bits, a set of
rotate instructions can be used. They move the bits that are shifted out of one end of the
operand back into the other end. Two versions of both the left and right rotate instructions

are usually provided. In one version, the bits of the operand are simply rotated. In the
other version, the rotation includes the C flag.
(a) Rotate left without carry RotateL #2, R0
om
C
R0
c
p.
Before: 0 0 1 1 1 0 . . . 0 1 1
ou
After: 1 1 1 0 . . . 0 1 1 0 1
gr
(b) Rotate left with carry
ts
RotateLC #2, R0
en
C R0
ud
st
0 1 1 1 0 . . . 0 11
Before: 0
ity
1 1 1 0 . . 0 1 1 0 0
.c
after:
w
w
w

(c ) Rotate right without carry RotateR #2, R0
om
R0
Before: 0 1 1 1 0 . . . 0 1 1 0
c
p.
1 1 0 1 1 1 0 . . . 0 1
ou
After:
(d) Rotate right with carry RotateRC #2, R0
gr
ts
R0 C
en
ud
Before: 0 1 1 1 0 . . . 0 1 1 0
st
after: 1 0 0 1 1 1 0 . . . 0 1
ity
.c
2.7 Encoding of machine instructions

We have introduced a variety of useful instructions and addressing modes. These
w
instructions specify the actions that must be performed by the processor circuitry to carry
out the desired tasks. We have often referred to them as machine instructions. Actually,
w
the form in which we have presented the instructions is indicative of the form used in
assembly languages, except that we tried to avoid using acronyms for the various
w
operations, which are awkward to memorize and are likely to be specific to a particular
commercial processor. To be executed in a processor, an instruction must be encoded in a
compact binary pattern. Such encoded instructions are properly referred to as machine
instructions. The instructions that use symbolic names and acronyms are called assembly

language instructions, which are converted into the machine instructions using the
assembler program.
We have seen instructions that perform operations such as add, subtract, move,
shift, rotate, and branch. These instructions may use operands of different sizes, such as
32-bit and 8-bit numbers or 8-bit ASCII-encoded characters. The type of operation that is
om
to be performed and the type of operands used may be specified using an encoded binary
pattern referred to as the OP code for the given instruction. Suppose that 8 bits are
allocated for this purpose, giving 256 possibilities for specifying different instructions.
c
This leaves 24 bits to specify the rest of the required information.
p.
Let us examine some typical cases. The instruction
Add R1, R2
ou
Has to specify the registers R1 and R2, in addition to the OP code. If the processor has 16
registers, then four bits are needed to identify each register. Additional bits are needed to
gr
indicate that the Register addressing mode is used for each operand.
The instruction ts
Move 24(R0), R5
en
Requires 16 bits to denote the OP code and the two registers, and some bits to express
that the source operand uses the Index addressing mode and that the index value is 24.
The shift instruction
ud
LShiftR #2, R0
And the move instruction

st
Move #$3A, R1
Have to indicate the immediate values 2 and #$3A, respectively, in addition to the 18
ity
bits used to specify the OP code, the addressing modes, and the register. This limits the
size of the immediate operand to what is expressible in 14 bits.
.c
Consider next the branch instruction

Branch >0 LOOP
w
Again, 8 bits are used for the OP code, leaving 24 bits to specify the branch
w
offset. Since the offset is a 2s-complement number, the branch target address must be
within 223 bytes of the location of the branch instruction. To branch to an instruction
w
outside this range, a different addressing mode has to be used, such as Absolute or
Register Indirect. Branch instructions that use these modes are usually called Jump
instructions.

In all these examples, the instructions can be encoded in a 32-bit word. Depicts a
possible format. There is an 8-bit Op-code field and two 7-bit fields for specifying the
source and destination operands. The 7-bit field identifies the addressing mode and the
register involved (if any). The Other info field allows us to specify the additional
information that may be needed, such as an index value or an immediate operand.
om
But, what happens if we want to specify a memory operand using the Absolute
addressing mode? The instruction
c
Move R2, LOC
p.
(a) One-word instruction
ou
Opcode Source Dest Other info
gr
(b) Two-Word instruction
ts
Opcode Source Dest Other info
en
Memory address/Immediate operand

ud
(c ) Three-operand instruction
st
ity
Op code Ri Rj Rk Other info

.c
Requires 18 bits to denote the OP code, the addressing modes, and the register.
This leaves 14 bits to express the address that corresponds to LOC, which is clearly
w
insufficient.
w
And #$FF000000. R2
w
In which case the second word gives a full 32-bit immediate operand.
If we want to allow an instruction in which two operands can be specified using

the Absolute addressing mode, for example

Move LOC1, LOC2
Then it becomes necessary to use tow additional words for the 32-bit addresses of
the operands.
This approach results in instructions of variable length, dependent on the number
om
of operands and the type of addressing modes used. Using multiple words, we can
implement quite complex instructions, closely resembling operations in high-level
programming languages. The term complex instruction set computer (CISC) has been
c
used to refer to processors that use instruction sets of this type.
p.
The restriction that an instruction must occupy only one word has led to a style of
computers that have become known as reduced instruction set computer (RISC). The
ou
RISC approach introduced other restrictions, such as that all manipulation of data must be
done on operands that are already in processor registers. This restriction means that the
above addition would need a two-instruction sequence
gr
Move (R3), R1
ts
Add R1, R2
en
If the Add instruction only has to specify the two registers, it will need just a
portion of a 32-bit word. So, we may provide a more powerful instruction that uses three
operands
ud
Add R1, R2, R3

st
Which performs the operation

ity
R3 [R1] + [R2]
.c
A possible format for such an instruction in shown in fig c. Of course, the

processor has to be able to deal with such three-operand instructions. In an instruction set
w
where all arithmetic and logical operations use only register operands, the only memory
references are made to load/store the operands into/from the processor registers.
w
RISC-type instruction sets typically have fewer and less complex instructions
than CISC-type sets. We will discuss the relative merits of RISC and CISC approaches in
w
Chapter 8, which deals with the details of processor design.

UNIT - 3
Input/output Organization: Accessing I/O Devices, Interrupts Interrupt
Hardware, Enabling and Disabling Interrupts, Handling Multiple Devices,
om
Controlling Device Requests, Exceptions, Direct Memory Access, Buses
c
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
w

CHAPTER 03
THE SYSTEM MEMORY
3.1 ACCESSING I/O DEVICES
om
A simple arrangement to connect I/O devices to a computer is to use a single bus
arrangement. The bus enables all the devices connected to it to exchange information.
c
Typically, it consists of three sets of lines used to carry address, data, and control signals.
p.
Each I/O device is assigned a unique set of addresses. When the processor places a
particular address on the address line, the device that recognizes this address responds to
ou
the commands issued on the control lines. The processor requests either a read or a write
operation, and the requested data are transferred over the data lines, when I/O devices and
gr
the memory share the same address space, the arrangement is called memory-mapped
I/O. ts
With memory-mapped I/O, any machine instruction that can access memory can
en
be used to transfer data to or from an I/O device. For example, if DATAIN is the address
of the input buffer associated with the keyboard, the instruction
ud
Move DATAIN, R0
Reads the data from DATAIN and stores them into processor register R0. Similarly, the
st
instruction
ity
Move R0, DATAOUT

.c
Sends the contents of register R0 to location DATAOUT, which may be the output data
buffer of a display unit or a printer.
w
Most computer systems use memory-mapped I/O. some processors have special
w
In and Out instructions to perform I/O transfers. When building a computer system based
on these processors, the designer had the option of connecting I/O devices to use the
w
special I/O address space or simply incorporating them as part of the memory address
space. The I/O devices examine the low-order bits of the address bus to determine
whether they should respond.

The hardware required to connect an I/O device to the bus. The address decoder
enables the device to recognize its address when this address appears on the address lines.
The data register holds the data being transferred to or from the processor. The status
register contains information relevant to the operation of the I/O device. Both the data
om
and status registers are connected to the data bus and assigned unique addresses. The
address decoder, the data and status registers, and the control circuitry required to
coordinate I/O transfers constitute the devices interface circuit.
c
p.
I/O devices operate at speeds that are vastly different from that of the processor.
When a human operator is entering characters at a keyboard, the processor is capable of
ou
executing millions of instructions between successive character entries. An instruction
that reads a character from the keyboard should be executed only when a character is
gr
available in the input buffer of the keyboard interface. Also, we must make sure that an
input character is read only once.
ts
This example illustrates program-controlled I/O, in which the processor
en
repeatedly checks a status flag to achieve the required synchronization between the
processor and an input or output device. We say that the processor polls the device. There
are two other commonly used mechanisms for implementing I/O operations: interrupts
ud
and direct memory access. In the case of interrupts, synchronization is achieved by

having the I/O device send a special signal over the bus whenever it is ready for a data
st
transfer operation. Direct memory access is a technique used for high-speed I/O devices.
ity
It involves having the device interface transfer data directly to or from the memory,
without continuous involvement by the processor.
.c
The routine executed in response to an interrupt request is called the interrupt-

service routine, which is the PRINT routine in our example. Interrupts bear considerable
w
resemblance to subroutine calls. Assume that an interrupt request arrives during

w
execution of instruction i in figure 1

w

Program 1 Program 2
COMPUTER routine PRINT routine
c om
1
p.
2
ou
....
Interrupt i
gr
Occurs i+1
here ts
en
M
ud
Figure 1. Transfer of control through the use of interrupts

st
The processor first completes execution of instruction i. Then, it loads the

ity
program counter with the address of the first instruction of the interrupt-service routine.
For the time being, let us assume that this address is hardwired in the processor. After
.c
execution of the interrupt-service routine, the processor has to come back to instruction
i +1. Therefore, when an interrupt occurs, the current contents of the PC, which point to
w
instruction i+1, must be put in temporary storage in a known location. A Return-from-

w
interrupt instruction at the end of the interrupt-service routine reloads the PC from the
temporary storage location, causing execution to resume at instruction i +1. In many
w
processors, the return address is saved on the processor stack.

We should note that as part of handling interrupts, the processor must inform the
device that its request has been recognized so that it may remove its interrupt-request

signal. This may be accomplished by means of a special control signal on the bus. An
interrupt-acknowledge signal. The execution of an instruction in the interrupt-service
routine that accesses a status or data register in the device interface implicitly informs
that device that its interrupt request has been recognized.
So far, treatment of an interrupt-service routine is very similar to that of a
om
subroutine. An important departure from this similarity should be noted. A subroutine
performs a function required by the program from which it is called. However, the
c
interrupt-service routine may not have anything in common with the program being
p.
executed at the time the interrupt request is received. In fact, the two programs often
belong to different users. Therefore, before starting execution of the interrupt-service
ou
routine, any information that may be altered during the execution of that routine must be
saved. This information must be restored before execution of the interrupt program is
gr
resumed. In this way, the original program can continue execution without being affected
in any way by the interruption, except for the time delay. The information that needs to
ts
be saved and restored typically includes the condition code flags and the contents of any
en
registers used by both the interrupted program and the interrupt-service routine.
The task of saving and restoring information can be done automatically by the
processor or by program instructions. Most modern processors save only the minimum
ud
amount of information needed to maintain the registers involves memory transfers that
increase the total execution time, and hence represent execution overhead. Saving
st
registers also increase the delay between the time an interrupt request is received and the
ity
start of execution of the interrupt-service routine. This delay is called interrupt latency.
.c
3.2 INTERRUPT HARDWARE:

We pointed out that an I/O device requests an interrupt by activating a bus line
w
called interrupt-request. Most computers are likely to have several I/O devices that can
w
request an interrupt. A single interrupt-request line may be used to serve n devices as

w
depicted. All devices are connected to the line via switches to ground. To request an
interrupt, a device closes its associated switch. Thus, if all interrupt-request signals
INTR1 to INTRn are inactive, that is, if all switches are open, the voltage on the interrupt-
request line will be equal to Vdd. This is the inactive state of the line. Since the closing of

one or more switches will cause the line voltage to drop to 0, the value of INTR is the
logical OR of the requests from individual devices, that is,
INTR = INTR1 + +INTRn
om
It is customary to use the complemented form, INTR , to name the interrupt-request
signal on the common line, because this signal is active when in the low-voltage state.
c
p.
3.3 ENABLING AND DISABLING INTERRUPTS:
The facilities provided in a computer must give the programmer complete control
ou
over the events that take place during program execution. The arrival of an interrupt
request from an external device causes the processor to suspend the execution of one
gr
program and start the execution of another. Because interrupts can arrive at any time,
they may alter the sequence of events from the envisaged by the programmer. Hence, the
ts
interruption of program execution must be carefully controlled.
en
Let us consider in detail the specific case of a single interrupt request from one
device. When a device activates the interrupt-request signal, it keeps this signal activated
ud
until it learns that the processor has accepted its request. This means that the interrupt-
request signal will be active during execution of the interrupt-service routine, perhaps
until an instruction is reached that accesses the device in question.
st
The first possibility is to have the processor hardware ignore the interrupt-request
ity
line until the execution of the first instruction of the interrupt-service routine has been
completed. Then, by using an Interrupt-disable instruction as the first instruction in the
.c
interrupt-service routine, the programmer can ensure that no further interruptions will
occur until an Interrupt-enable instruction is executed. Typically, the Interrupt-enable
w
instruction will be the last instruction in the interrupt-service routine before the Return-
w
from-interrupt instruction. The processor must guarantee that execution of the Return-
w
from-interrupt instruction is completed before further interruption can occur.

The second option, which is suitable for a simple processor with only one
interrupt-request line, is to have the processor automatically disable interrupts before
starting the execution of the interrupt-service routine. After saving the contents of the PC

and the processor status register (PS) on the stack, the processor performs the equivalent
of executing an Interrupt-disable instruction. It is often the case that one bit in the PS
register, called Interrupt-enable, indicates whether interrupts are enabled.
In the third option, the processor has a special interrupt-request line for which the
interrupt-handling circuit responds only to the leading edge of the signal. Such a line is
om
said to be edge-triggered.
Before proceeding to study more complex aspects of interrupts, let us summarize
c
the sequence of events involved in handling an interrupt request from a single device.
p.
Assuming that interrupts are enabled, the following is a typical scenario.
1. The device raises an interrupt request.
ou
2. The processor interrupts the program currently being executed.
3. Interrupts are disabled by changing the control bits in the PS (except in the case of
gr
edge-triggered interrupts).
4. The device is informed that its request has been recognized, and in response, it
ts
deactivates the interrupt-request signal.
en
5. The action requested by the interrupt is performed by the interrupt-service routine.
6. Interrupts are enabled and execution of the interrupted program is resumed.
ud
3.4 HANDLING MULTIPLE DEVICES:

Let us now consider the situation where a number of devices capable of initiating
st
interrupts are connected to the processor. Because these devices are operationally
ity
independent, there is no definite order in which they will generate interrupts. For
example, device X may request in interrupt while an interrupt caused by device Y is
.c
being serviced, or several devices may request interrupts at exactly the same time. This
gives rise to a number of questions
w
w
1. How can the processor recognize the device requesting an interrupts?

2. Given that different devices are likely to require different interrupt-service
w
routines, how can the processor obtain the starting address of the appropriate
routine in each case?

3. Should a device be allowed to interrupt the processor while another interrupt is

being serviced?
4. How should two or more simultaneous interrupt requests be handled?
The means by which these problems are resolved vary from one computer to another,
And the approach taken is an important consideration in determining the computers
om
suitability for a given application.
When a request is received over the common interrupt-request line, additional
c
information is needed to identify the particular device that activated the line.
p.
The information needed to determine whether a device is requesting an interrupt
is available in its status register. When a device raises an interrupt request, it sets to 1 one
ou
of the bits in its status register, which we will call the IRQ bit. For example, bits KIRQ
and DIRQ are the interrupt request bits for the keyboard and the display, respectively.
gr
The simplest way to identify the interrupting device is to have the interrupt-service
routine poll all the I/O devices connected to the bus. The first device encountered with its
ts
IRQ bit set is the device that should be serviced. An appropriate subroutine is called to
en
provide the requested service.
The polling scheme is easy to implement. Its main disadvantage is the time spent
interrogating the IRQ bits of all the devices that may not be requesting any service. An
ud
alternative approach is to use vectored interrupts, which we describe next.

st
Vectored Interrupts:-
ity
To reduce the time involved in the polling process, a device requesting an

.c
interrupt may identify itself directly to the processor. Then, the processor can
immediately start executing the corresponding interrupt-service routine. The term
w
vectored interrupts refers to all interrupt-handling schemes based on this approach.

w
A device requesting an interrupt can identify itself by sending a special code to

w
the processor over the bus. This enables the processor to identify individual devices even
if they share a single interrupt-request line. The code supplied by the device may
represent the starting address of the interrupt-service routine for that device. The code

length is typically in the range of 4 to 8 bits. The remainder of the address is supplied by
the processor based on the area in its memory where the addresses for interrupt-service
routines are located.
This arrangement implies that the interrupt-service routine for a given device
om
must always start at the same location. The programmer can gain some flexibility by
storing in this location an instruction that causes a branch to the appropriate routine.
c
p.
Interrupt Nesting: -
ou
Interrupts should be disabled during the execution of an interrupt-service routine,
to ensure that a request from one device will not cause more than one interruption. The
gr
same arrangement is often used when several devices are involved, in which case
execution of a given interrupt-service routine, once started, always continues to
ts
completion before the processor accepts an interrupt request from a second device.
en
Interrupt-service routines are typically short, and the delay they may cause is acceptable
for most simple devices.
ud
For some devices, however, a long delay in responding to an interrupt request

may lead to erroneous operation. Consider, for example, a computer that keeps track of
st
the time of day using a real-time clock. This is a device that sends interrupt requests to
ity
the processor at regular intervals. For each of these requests, the processor executes a
short interrupt-service routine to increment a set of counters in the memory that keep
.c
track of time in seconds, minutes, and so on. Proper operation requires that the delay in
responding to an interrupt request from the real-time clock be small in comparison with
w
the interval between two successive requests. To ensure that this requirement is satisfied
w
in the presence of other interrupting devices, it may be necessary to accept an interrupt

request from the clock during the execution of an interrupt-service routine for another
w
device.

This example suggests that I/O devices should be organized in a priority

structure. An interrupt request from a high-priority device should be accepted while the
processor is servicing another request from a lower-priority device.
A multiple-level priority organization means that during execution of an
om
interrupt-service routine, interrupt requests will be accepted from some devices but not
from others, depending upon the devices priority. To implement this scheme, we can
c
assign a priority level to the processor that can be changed under program control. The
p.
priority level of the processor is the priority of the program that is currently being
executed. The processor accepts interrupts only from devices that have priorities higher
ou
than its own.
gr
The processors priority is usually encoded in a few bits of the processor status
word. It can be changed by program instructions that write into the PS. These are
ts
privileged instructions, which can be executed only while the processor is running in the
en
supervisor mode. The processor is in the supervisor mode only when executing operating
system routines. It switches to the user mode before beginning to execute application
programs. Thus, a user program cannot accidentally, or intentionally, change the priority
ud
of the processor and disrupt the systems operation. An attempt to execute a privileged
instruction while in the user mode leads to a special type of interrupt called a privileged
st
instruction.
ity
A multiple-priority scheme can be implemented easily by using separate

.c
interrupt-request and interrupt-acknowledge lines for each device, as shown in figure.

Each of the interrupt-request lines is assigned a different priority level. Interrupt requests
w
received over these lines are sent to a priority arbitration circuit in the processor. A
w
request is accepted only if it has a higher priority level than that currently assigned to the
processor.
w

INTR p
om
INTR 1
Device 1 Device 2 Device p
c
p.
INTA 1 INTA p
ou
Priority arbitration Circuit
Figure2: Implementation of interrupt priority using individual interrupt-request and
gr
acknowledge lines.
Simultaneous Requests:-
ts
en
Let us now consider the problem of simultaneous arrivals of interrupt requests
from two or more devices. The processor must have some means of deciding which
ud
requests to service first. Using a priority scheme such as that of figure, the solution is
straightforward. The processor simply accepts the requests having the highest priority.
st
Polling the status registers of the I/O devices is the simplest such mechanism. In
ity
this case, priority is determined by the order in which the devices are polled. When
vectored interrupts are used, we must ensure that only one device is selected to send its
.c
interrupt vector code. A widely used scheme is to connect the devices to form a daisy
w
chain, as shown in figure 3a. The interrupt-request line INTR is common to all devices.
w
The interrupt-acknowledge line, INTA, is connected in a daisy-chain fashion, such that

the INTA signal propagates serially through the devices.
w

INTR
om
Device 1 Device 2 Device n
c
INTA
p.
(3.a) Daisy chain
ou
gr
INTR 1
ts
en
Device Device
INTA1
ud
INTR p
st
ity
Device Device
.c
INTA p
w
Priority arbitration
w
Circuit (3.b) Arrangement of priority groups

w
When several devices raise an interrupt request and the INTR line is activated,
the processor responds by setting the INTA line to 1. This signal is received by device 1.

Device 1 passes the signal on to device 2 only if it does not require any service. If device
1 has a pending request for interrupt, it blocks the INTA signal and proceeds to put its
identifying code on the data lines. Therefore, in the daisy-chain arrangement, the device
that is electrically closest to the processor has the highest priority. The second device
along the chain has second highest priority, and so on.
om
The scheme in figure 3.a requires considerably fewer wires than the individual
c
connections in figure 2. The main advantage of the scheme in figure 2 is that it allows the
p.
processor to accept interrupt requests from some devices but not from others, depending
upon their priorities. The two schemes may be combined to produce the more general
ou
structure in figure 3b. Devices are organized in groups, and each group is connected at a
different priority level. Within a group, devices are connected in a daisy chain. This
gr
organization is used in many computer systems.
ts
3.5 CONTROLLING DEVICE REQUESTS:
en
Until now, we have assumed that an I/O device interface generates an interrupt
ud
request whenever it is ready for an I/O transfer, for example whenever the SIN flag is 1.
It is important to ensure that interrupt requests are generated only by those I/O devices
that are being used by a given program. Idle devices must not be allowed to generate
st
interrupt requests, even though they may be ready to participate in I/O transfer
ity
operations. Hence, we need a mechanism in the interface circuits of individual devices to

control whether a device is allowed to generate an interrupt request.
.c
The control needed is usually provided in the form of an interrupt-enable bit in

w
the devices interface circuit. The keyboard interrupt-enable, KEN, and display interrupt-
w
enable, DEN, flags in register CONTROL perform this function. If either of these flags
is set, the interface circuit generates an interrupt request whenever the corresponding
w
status flag in register STATUS is set. At the same time, the interface circuit sets bit KIRQ
or DIRQ to indicate that the keyboard or display unit, respectively, is requesting an

interrupt. If an interrupt-enable bit is equal to 0, the interface circuit will not generate an
interrupt request, regardless of the state of the status flag.
To summarize, there are two independent mechanisms for controlling interrupt

requests. At the device end, an interrupt-enable bit in a control register determines
om
whether the device is allowed to generate an interrupt request. At the processor end,
either an interrupt enable bit in the PS register or a priority structure determines whether
c
a given interrupt request will be accepted.
p.
3.6 EXCEPTIONS:
ou
An interrupt is an event that causes the execution of one program to be suspended
and the execution of another program to begin. So far, we have dealt only with interrupts
gr
caused by requests received during I/O data transfers. However, the interrupt mechanism
is used in a number of other situations.
ts
en
The term exception is often used to refer to any event that causes an interruption.
Hence, I/O interrupts are one example of an exception. We now describe a few other
ud
kinds of exceptions.
Recovery from Errors:

st
ity
Computers use a variety of techniques to ensure that all hardware components are
operating properly. For example, many computers include an error-checking code in the
.c
main memory, which allows detection of errors in the stored data. If errors occur, the
control hardware detects it and informs the processor by raising an interrupt.
w
w
The processor may also interrupt a program if it detects an error or an unusual

condition while executing the instructions of this program. For example, the OP-code
w
field of an instruction may not correspond to any legal instruction, or an arithmetic

instruction may attempt a division by zero.

When exception processing is initiated as a result of such errors, the processor

proceeds in exactly the same manner as in the case of an I/O interrupt request. It suspends
the program being executed and starts an exception-service routine. This routine takes
appropriate action to recover from the error, if possible, or to inform the user about it.
Recall that in the case of an I/O interrupt, the processor completes execution of the
om
instruction in progress before accepting the interrupt. However, when an interrupt is
caused by an error, execution of the interrupted instruction cannot usually be completed,
c
and the processor begins exception processing immediately.
p.
Debugging:
ou
Another important type of exception is used as an aid in debugging programs.
System software usually includes a program called a debugger, which helps the
gr
programmer find errors in a program. The debugger uses exceptions to provide two
important facilities called trace and breakpoints.
ts
en
When a processor is operating in the trace mode, an exception occurs after
execution of every instruction, using the debugging program as the exception-service
routine. The debugging program enables the user to examine the contents of registers,
ud
memory locations, and so on. On return from the debugging program, the next instruction
in the program being debugged is executed, then the debugging program is activated
st
again. The trace exception is disabled during the execution of the debugging program.
ity
Breakpoint provides a similar facility, except that the program being debugged is
.c
interrupted only at specific points selected by the user. An instruction called Trap or
Software-interrupt is usually provided for this purpose. Execution of this instruction
w
results in exactly the same actions as when a hardware interrupt request is received.
w
While debugging a program, the user may wish to interrupt program execution after
instruction i. The debugging routine saves instruction i+1 and replaces it with a software
w
interrupt instruction. When the program is executed and reaches that point, it is
interrupted and the debugging routine is activated. This gives the user a chance to
examine memory and register contents. When the user is ready to continue executing the

program being debugged, the debugging routine restores the saved instruction that was a
location i+1 and executes a Return-from-interrupt instruction.
Privilege Exception:
To protect the operating system of a computer from being corrupted by user
om
programs, certain instructions can be executed only while the processor is in supervisor
mode. These are called privileged instructions. For example, when the processor is
c
running in the user mode, it will not execute an instruction that changes the priority level
p.
of the processor or that enables a user program to access areas in the computer memory
that have been allocated to other users. An attempt to execute such an instruction will
ou
produce a privilege exceptions, causing the processor to switch to the supervisor mode
and begin executing an appropriate routine in the operating system.
gr
3.7 DIRECT MEMORY ACCESS: ts
The discussion in the previous sections concentrates on data transfer between the
en
processor and I/O devices. Data are transferred by executing instructions such as
ud
Move DATAIN, R0
An instruction to transfer input or output data is executed only after the processor
st
determines that the I/O device is ready. To do this, the processor either polls a status flag
ity
in the device interface or waits for the device to send an interrupt request. In either case,
considerable overhead is incurred, because several program instructions must be executed
.c
for each data word transferred. In addition to polling the status register of the device,
instructions are needed for incrementing the memory address and keeping track of the
w
word count. When interrupts are used, there is the additional overhead associated with
w
saving and restoring the program counter and other state information.
w
To transfer large blocks of data at high speed, an alternative approach is used. A

special control unit may be provided to allow transfer of a block of data directly between

an external device and the main memory, without continuous intervention by the
processor. This approach is called direct memory access, or DMA.
DMA transfers are performed by a control circuit that is part of the I/O device
interface. We refer to this circuit as a DMA controller. The DMA controller performs the
om
functions that would normally be carried out by the processor when accessing the main
memory. For each word transferred, it provides the memory address and all the bus
c
signals that control data transfer. Since it has to transfer blocks of data, the DMA
p.
controller must increment the memory address for successive words and keep track of the
number of transfers.
ou
Although a DMA controller can transfer data without intervention by the
gr
processor, its operation must be under the control of a program executed by the
processor. To initiate the transfer of a block of words, the processor sends the starting
ts
address, the number of words in the block, and the direction of the transfer. On receiving
en
this information, the DMA controller proceeds to perform the requested operation. When
the entire block has been transferred, the controller informs the processor by raising an
interrupt signal.
ud
While a DMA transfer is taking place, the program that requested the transfer
st
cannot continue, and the processor can be used to execute another program. After the
ity
DMA transfer is completed, the processor can return to the program that requested the
transfer.
.c
I/O operations are always performed by the operating system of the computer in
w
response to a request from an application program. The OS is also responsible for

w
suspending the execution of one program and starting another. Thus, for an I/O operation
involving DMA, the OS puts the program that requested the transfer in the Blocked state,
w
initiates the DMA operation, and starts the execution of another program. When the
transfer is completed, the DMA controller informs the processor by sending an interrupt

request. In response, the OS puts the suspended program in the Runnable state so that it
can be selected by the scheduler to continue execution.
Figure 4 shows an example of the DMA controller registers that are accessed by
the processor to initiate transfer operations. Two registers are used for storing the
om
Status and Control 31 30 1 0
c
p.
IRQ Done
ou
IE R/W
gr
Starting address
ts
en
Word count
Figure 4 Registers in DMA interface

ud
Main
memory
st
Processor
ity
.c
System bus
w
w
Disk/DMA DMA
controller controller Printer Keyboard
w

Disk Disk Network

Interface
om
Figure 5 Use of DMA controllers in a computer system
c
p.
Starting address and the word count. The third register contains status and control flags.
The R/W bit determines the direction of the transfer. When this bit is set to 1 by a
ou
program instruction, the controller performs a read operation, that is, it transfers data
from the memory to the I/O device. Otherwise, it performs a write operation. When the
gr
controller has completed transferring a block of data and is ready to receive another
command, it sets the Done flag to 1. Bit 30 is the Interrupt-enable flag, IE. When this flag
ts
is set to 1, it causes the controller to raise an interrupt after it has completed transferring a
en
block of data. Finally, the controller sets the IRQ bit to 1 when it has requested an
interrupt.
ud
An example of a computer system is given in above figure, showing how DMA

controllers may be used. A DMA controller connects a high-speed network to the
st
computer bus. The disk controller, which controls two disks, also has DMA capability
ity
and provides two DMA channels. It can perform two independent DMA operations, as if
each disk had its own DMA controller. The registers needed to store the memory address,
.c
the word count, and so on are duplicated, so that one set can be used with each device.
w
To start a DMA transfer of a block of data from the main memory to one of the
w
disks, a program writes the address and word count information into the registers of the
corresponding channel of the disk controller. It also provides the disk controller with
w
information to identify the data for future retrieval. The DMA controller proceeds
independently to implement the specified operation. When the DMA transfer is
completed. This fact is recorded in the status and control register of the DMA channel by

setting the Done bit. At the same time, if the IE bit is set, the controller sends an interrupt
request to the processor and sets the IRQ bit. The status register can also be used to
record other information, such as whether the transfer took place correctly or errors
occurred.
Memory accesses by the processor and the DMA controller are interwoven.
om
Requests by DMA devices for using the bus are always given higher priority than
processor requests. Among different DMA devices, top priority is given to high-speed
c
peripherals such as a disk, a high-speed network interface, or a graphics display device.
p.
Since the processor originates most memory access cycles, the DMA controller can be
said to steal memory cycles from the processor. Hence, the interweaving technique is
ou
usually called cycle stealing. Alternatively, the DMA controller may be given exclusive
access to the main memory to transfer a block of data without interruption. This is known
gr
as block or burst mode.
ts
Most DMA controllers incorporate a data storage buffer. In the case of the
en
network interface in figure 5 for example, the DMA controller reads a block of data from
the main memory and stores it into its input buffer. This transfer takes place using burst
mode at a speed appropriate to the memory and the computer bus. Then, the data in the
ud
buffer are transmitted over the network at the speed of the network.
st
A conflict may arise if both the processor and a DMA controller or two DMA
ity
controllers try to use the bus at the same time to access the main memory. To resolve
these conflicts, an arbitration procedure is implemented on the bus to coordinate the
.c
activities of all devices requesting memory transfers.

w
Bus Arbitration:-
w
The device that is allowed to initiate data transfers on the bus at any given time is
w
called the bus master. When the current master relinquishes control of the bus, another
device can acquire this status. Bus arbitration is the process by which the next device to
become the bus master is selected and bus mastership is transferred to it. The selection of

the bus master must take into account the needs of various devices by establishing a
priority system for gaining access to the bus.
There are two approaches to bus arbitration: centralized and distributed. In

centralized arbitration, a single bus arbiter performs the required arbitration. In
om
distributed arbitration, all devices participate in the selection of the next bus master.
c
p.
Centralized Arbitration:-
ou
The bus arbiter may be the processor or a separate unit connected to the bus. A
basic arrangement in which the processor contains the bus arbitration circuitry. In this
gr
case, the processor is normally the bus master unless it grants bus mastership to one of
the DMA controllers. A DMA controller indicates that it needs to become the bus master
ts
by activating the Bus-Request line, BR . The signal on the Bus-Request line is the logical
en
OR of the bus requests from all the devices connected to it. When Bus-Request is
activated, the processor activates the Bus-Grant signal, BG1, indicating to the DMA
controllers that they may use the bus when it becomes free. This signal is connected to all
ud
DMA controllers using a daisy-chain arrangement. Thus, if DMA controller 1 is

requesting the bus, it blocks the propagation of the grant signal to other devices.
st
Otherwise, it passes the grant downstream by asserting BG2. The current bus master
ity
indicates to all device that it is using the bus by activating another open-controller line
called Bus-Busy, BBSY . Hence, after receiving the Bus-Grant signal, a DMA controller
.c
waits for Bus-Busy to become inactive, then assumes mastership of the bus. At this time,
it activates Bus-Busy to prevent other devices from using the bus at the same time.
w
w
w

Distributed Arbitration:-
Vcc ARB3
om
ARB2
c
ARB1
p.
ou
ARB0
Start-Arbitration
gr
ts
en
O.C
ud
0 1 0 1 0 1 1 1
st
Interface circuit
for device A
ity
.c
Fig 6 A distributed arbitration

w
w
Distributed arbitration means that all devices waiting to use the bus have equal
responsibility in carrying out the arbitration process, without using a central arbiter. A
w
simple method for distributed arbitration is illustrated in figure 6. Each device on the bus
assigned a 4-bit identification number. When one or more devices request the bus, they
assert the StartArbitration signal and place their 4-bit ID numbers on four open-
Dept of Page 80
CSE,SJBIT
collector lines, ARB 0 through ARB 3 . A winner is selected as a result of the

interaction among the signals transmitted over those liens by all contenders. The net
outcome is that the code on the four lines represents the request that has the highest ID
number.
om
3.8 BUSES:
The processor, main memory, and I/O devices can be interconnected by means of
c
a common bus whose primary function is to provide a communication path for the
p.
transfer of data. The bus includes the lines needed to support interrupts and arbitration. In
this section, we discuss the main features of the bus protocols used for transferring data.
ou
A bus protocol is the set of rules that govern the behavior of various devices connected to
the bus as to when to place information on the bus, assert control signals, and so on. After
gr
describing bus protocols, we will present examples of interface circuits that use these
protocols. ts
Synchronous Bus:-
en
In a synchronous bus, all devices derive timing information from a common

ud
clock line. Equally spaced pulses on this line define equal time intervals. In the simplest
form of a synchronous bus, each of these intervals constitutes a bus cycle during which
st
one data transfer can take place. Such a scheme is illustrated in figure 7 The address and
ity
data lines in this and subsequent figures are shown as high and low at the same time. This
is a common convention indicating that some lines are high and some low, depending on
.c
the particular address or data pattern being transmitted. The crossing points indicate the
times at which these patterns change. A signal line in an indeterminate or high impedance
w
state is represented by an intermediate level half-way between the low and high signal
w
levels.
Let us consider the sequence of events during an input (read) operation. At time
w
t0, the master places the device address on the address lines and sends an appropriate
command on the control lines. In this case, the command will indicate an input operation
and specify the length of the operand to be read, if necessary. Information travels over the

bus at a speed determined by its physical and electrical characteristics. The clock pulse
width, t1 t0, must be longer than the maximum propagation delay between two devices
connected to the bus. It also has to be long enough to allow all devices to decode the
address and control signals so that the addressed device (the slave) can respond at time t1.
It is important that slaves take no action or place any data on the bus before t1. The
om
information on the bus is unreliable during the period t0 to t1 because signals are changing
state. The addressed slave places the requested input data on the data lines at time t1.
c
At the end of the clock cycle, at time t2, the master strobes the data on the data
p.
lines into its input buffer. In this context, strobe means to capture the values of the .
Figure 7 Timing of an input transfer on a synchronous bus.
ou
Time
gr
Bus clock
ts
en
Address
and
ud
command
st
Data
ity
.c
t1
t0 t2
w
w
Bus cycle
Data of a given instant and store them into a buffer. For data to be loaded correctly into
w
any storage device, such as a register built with flip-flops, the data must be available at
the input of that device for a period greater than the setup time of the device. Hence, the

period t2 - t1 must be greater than the maximum propagation time on the bus plus the
setup time of the input buffer register of the master.
A similar procedure is followed for an output operation. The master places the
output data on the data lines when it transmits the address and command information at
om
time t2, the addressed device strobes the data lines and loads the data into its data buffer.
c
The timing diagram in figure 7 is an idealized representation of the actions that
p.
take place on the bus lines. The exact times at which signals actually change state are
somewhat different from those shown because of propagation delays on bus wires and in
ou
the circuits of the devices. Figure 4.24 gives a more realistic picture of what happens in
practice. It shows two views of each signal, except the clock. Because signals take time to
gr
travel from one device to another, a given signal transition is seen by different devices at
different times. One view shows the signal as seen by the master and the other as seen by
ts
the slave.
en
The master sends the address and command signals on the rising edge at the
beginning of clock period 1 (t0). However, these signals do not actually appear on the bus
ud
until fAM, largely due to the delay in the bus driver circuit. A while later, at tAS, the
signals reach the slave. The slave decodes the address and at t1 sends the requested data.
st
Here again, the data signals do not appear on the bus until tDS. They travel toward the
ity
master and arrive at tDM. At t2, the master loads the data into its input buffer. Hence the
period t2-tDM is the setup time for the masters input buffer. The data must continue to be
.c
valid after t2 for a period equal to the hold time of that buffer.
w
Figure 8 A detailed timing diagram for the input transfer of figure 7

w
w

Time
Bus clock
om
Seen by master tAM
c
p.
Address
and
ou
command
gr
Data tDM
Seen by slave tAS

ts
en
Address
and
command
ud
Data tDS
st
ity
t0 t1 t2
.c
w
Multiple-Cycle transfers:-
w
The scheme described above results in a simple design for the device interface,
w
however, it has some limitations. Because a transfer has to be completed within one clock
cycle, the clock period, t2-t0, must be chosen to accommodate the longest delays on the

bus and the lowest device interface. This forces all devices to operate at the speed of the
slowest device.
Also, the processor has no way of determining whether the addressed device has
actually responded. It simply assumes that, at t2, the output data have been received by
om
the I/O device or the input data are available on the data lines. If, because of a
malfunction, the device does not respond, the error will not be detected.
c
p.
To overcome these limitations, most buses incorporate control signals that
represent a response from the device. These signals inform the master that the slave has
ou
recognized its address and that it is ready to participate in a data-transfer operation. They
also make it possible to adjust the duration of the data-transfer period to suit the needs of
gr
the participating devices. To simplify this process, a high-frequency clock signal is used
such that a complete data transfer cycle would span several clock cycles. Then, the
ts
number of clock cycles involved can vary from one device to another.
en
An example of this approach is shown in figure 4.25. during clock cycle 1, the
master sends address and command information on the bus, requesting a read operation.
ud
The slave receives this information and decodes it. On the following active edge of the
clock, that is, at the beginning of clock cycle 2, it makes a decision to respond and begins
st
to access the requested data. We have assumed that some delay is involved in getting the
ity
data, and hence the slave cannot respond immediately. The data become ready and are
placed on the bus in clock cycle 3. At the same time, the slave asserts a control signal
.c
called Slave-ready.
w
The Slave-ready signal is an acknowledgment from the slave to the master,

w
confirming that valid data have been sent. In the example in figure 9, the slave responds
in cycle 3. Another device may respond sooner or later. The Slave-ready signal allows the
w
duration of a bus transfer to change from one device to another. If the addressed device
does not respond at all, the master waits for some predefined maximum number of clock

cycles, then aborts the operation. This could be the result of an incorrect address or a
device malfunction.
Time
1 2 3 4
om
Clock
c
p.
Address
ou
Command
gr
Data
ts
en
ud
Slave-ready
st
ity
Figure 9 An input transfer using multiple clock cycles.
ASYNCHRONOUS BUS:-
.c
w
An alternative scheme for controlling data transfers on the bus is based on the use
of a handshake between the master and the salve. The concept of a handshake is a
w
generalization of the idea of the Slave-ready signal in figure 10. The common clock is
w
replaced by two timing control lines, Master-ready and Slave-ready. The first is asserted
by the master to indicate that it is ready for a transaction, and the second is a response
from the slave.

In principle, a data transfer controlled by a handshake protocol proceeds as

follows. The master places the address and command information on the bus. Then it
indicates to all devices that it has done so by activating the Master-ready line. This causes
all devices on the bus to decode the address. The selected slave performs the required
operation and informs the processor it has done so by activating the Slave-ready line. The
om
master waits for Slave-ready to become asserted before it removes its signals from the
bus. In the case of a read operation, it also strobes the data into its input buffer.
c
Figure 10 Handshake control of data transfer during an input operation.
p.
ou
Address
And command
gr
Master-ready
ts
en
ud
Slave-ready
st
ity
Data
.c
w
w
t0 t1 t2 t3 t4 t5
w
Bus cycle

An example of the timing of an input data transfer using the handshake scheme is
given in figure 4.26, which depicts the following sequence of events.
t0 The master places the address and command information on the bus, and all devices
on the bus begin to decode this information.
t1 The master sets the Master-ready line to 1 to inform the I/O devices that the address
om
and command information is ready. The delay t1-t0 is intended to allow for any skew that
may occur o the bus. Skew occurs when two signals simultaneously transmitted from one
c
source arrive at the destination at different times. This happens because different lines of
p.
the bus may have different propagation speeds. Thus, to guarantee that the Master-ready
signal does not arrive at any device ahead of the address and command information, the
ou
delay t1-t0 should be larger than the maximum possible bus skew.
t2 The selected slave, having decoded the address and command information performs
gr
the required input operation by placing the data from its data register on the data lines.
t3 The Slave-ready signal arrives at the master, indicating that the input data are
ts
available on the bus.
t4 The master removes the address and command information from the bus. The delay
en
between t3 and t4 is again intended to allow for bus skew.
t5 When the device interface receives the 1 to 0 transition of the Master-ready signal, it
ud
removes the data and the Slave-ready signal from the bus. This completes the input
transfer.
st
ity
.c
w
w
w

UNIT-4
Input/Output Organization contd.: Interface Circuits, Standard I/O
Interfaces PCI Bus, SCSI Bus, USB
c om
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
w

UNIT-4
INPUT/OUTPUT ORGANIZATION CONTD
4.1 INTERFACE CIRCUITS
om
Parallel port
c
The hardware components needed for connecting a keyboard to a processor. A
p.
typical keyboard consists of mechanical switches that are normally open. When a key is
pressed, its switch closes and establishes a path for an electrical signal. This signal is
ou
detected by an encoder circuit that generates the ASCII code for the corresponding
character.
gr
Figure 11 Keyboard to processor connection.
ts
en
Data Data
DATAIN
ud
Encoder
Address and Keyboard
Processor Debouncing switches
st
SIN circuit
ity
R/ W Input
interface
.c
Master-ready Valid
w
w
Slave-ready
w

The output of the encoder consists of the bits that represent the encoded character
and one control signal called Valid, which indicates that a key is being pressed. This
information is sent to the interface circuit, which contains a data register, DATAIN, and a
status flag, SIN. When a key is pressed, the Valid signal changes from 0 to 1, causing the
om
ASCII code to be loaded into DATAIN and SIN to be set to 1. The status flag SIN is
cleared to 0 when the processor reads the contents of the DATAIN register. The interface
c
circuit is connected to an asynchronous bus on which transfers are controlled using the
p.
handshake signals Master-ready and Slave-ready, as indicated in figure 11. The third
control line, R/ W distinguishes read and write transfers.
ou
Figure 12 shows a suitable circuit for an input interface. The output lines of the
gr
DATAIN register are connected to the data lines of the bus by means of three-state
drivers, which are turned on when the processor issues a read instruction with the address
ts
that selects this register. The SIN signal is generated by a status flag circuit. This signal is
en
also sent to the bus through a three-state driver. It is connected to bit D0, which means it
will appear as bit 0 of the status register. Other bits of this register do not contain valid
information. An address decoder is used to select the input interface when the high-order
ud
31 bits of an address correspond to any of the addresses assigned to this interface.

Address bit A0 determines whether the status or the data registers is to be read when the
st
Master-ready signal is active. The control handshake is accomplished by activating the

ity
Slave-ready signal when either Read-status or Read-data is equal to 1.

.c
w
w
w

c om
p.
ou
gr
tsFig 12 Input interface circuit
en
Data Data
ud
DATAOUT
Address Printer
st
Processor
SOUT
ity
Valid
R/ W
Output
.c
interface
Master-ready Idle
w
w
Slave-ready
w
Fig 13 Printer to processor connection
Dept of Page 92
CSE,SJBIT
Let us now consider an output interface that can be used to connect an output
device, such as a printer, to a processor, as shown in figure 13. The printer operates under
control of the handshake signals Valid and Idle in a manner similar to the handshake used
on the bus with the Master-ready and Slave-ready signals. When it is ready to accept a
character, the printer asserts its Idle signal. The interface circuit can then place a new
om
character on the data lines and activate the Valid signal. In response, the printer starts
printing the new character and negates the Idle signal, which in turn causes the interface
c
to deactivate the Valid signal.
p.
The circuit in figure 16 has separate input and output data lines for connection to
ou
an I/O device. A more flexible parallel port is created if the data lines to I/O devices are
bidirectional. Figure 17 shows a general-purpose parallel interface circuit that can be
gr
configured in a variety of ways. Data lines P7 through P0 can be used for either input or
output purposes. For increased flexibility, the circuit makes it possible for some lines to
ts
serve as inputs and some lines to serve as outputs, under program control. The
en
DATAOUT register is connected to these lines via three-state drivers that are controlled
by a data direction register, DDR. The processor can write any 8-bit pattern into DDR.
For a given bit, if the DDR value is 1, the corresponding data line acts as an output line;
ud
otherwise, it acts as an input line.

st
ity
.c
w
w
w

c om
p.
ou
gr
ts
Fig14 Output interface circuit
en
ud
st
ity
.c
w
w
w
Fig 15 combined input/output interface circuit

D7 P7
om
DATAIN
c
p.
D0 P0
ou
gr
ts
en
DATAOUT
ud
st
ity
.c
Data
Direction
Register
w
w
Fig 16 A general 8-bit interface

w

My-address
om
RS2 Register C1
Status
RS1 select and
Control
c
p.
RS0
R/W C2
ou
Ready
gr
Accept
INTR
ts
en
Fig 16 A general 8-bit parallel interface
ud
st
ity
.c
w
w
w

Fig 17 A parallel port interface for the bus
om
Fig 18 State diagram for the timing logic.
c
Time
p.
ou
1 2 3
gr
Clock
ts
en
Address
ud
R/ W
st
ity
.c
Data
w
w
Go
w
Slave-ready

Figure19 Timing for output interface
SERIAL PORT:-
A Serial port is used to connect the processor to I/O devices that require
om
transmission of data one bit at a time. The key feature of an interface circuit for a serial
port is that it is capable of communicating in a bit-serial fashion on the device side and in
c
a bit-parallel fashion on the bus side. The transformation between the parallel and serial
p.
formats is achieved with shift registers that have parallel access capability. A block
diagram of a typical serial interface is shown in figure 20. It includes the familiar
ou
DATAIN and DATAOUT registers. The input shift register accepts bit-serial input from
the I/O device. When all 8 bits of data have been received, the contents of this shift
gr
register are loaded in parallel into the DATAIN register. Similarly, output data in the
DATAOUT register are loaded into the output register, from which the bits are shifted
ts
out and sent to the I/O device.
en
ud
st
ity
.c
w
w
w

Figure 20 A serial interface
Input shift register

Serial
input
om
DATAIN
c
p.
ou
gr
ts
en
D7
ud
D0
st
DATAOUT
ity
My-address
.c
Chip and
RS1
Register select
w
w
RS0 Serial
Output shift register
R/ W output
w
Ready

Accept
Receiving clock
Status
om
and
INTR
Control
Transmission clock
c
p.
The double buffering used in the input and output paths are important. A simpler
interface could be implemented by turning DATAIN and DATAOUT into shift registers
ou
and eliminating the shift registers in figure 4.37. However, this would impose awkward
restrictions on the operation of the I/O device; after receiving one character from the
gr
serial line, the device cannot start receiving the next character until the processor reads
the contents of DATAIN. Thus, a pause would be needed between two characters to
ts
allow the processor to read the input data. With the double buffer, the transfer of the
en
second character can begin as soon as the first character is loaded from the shift register
into the DATAIN register. Thus, provided the processor reads the contents of DATAIN
before the serial transfer of the second character is completed, the interface can receive a
ud
continuous stream of serial data. An analogous situation occurs in the output path of the
interface.
st
ity
Because serial interfaces play a vital role in connecting I/O devices,

several widely used standards have been developed. A standard circuit that includes the
.c
features of our example in figure 20. is known as a Universal Asynchronous Receiver

Transmitter (UART). It is intended for use with low-speed serial devices. Data
w
transmission is performed using the asynchronous start-stop format. To facilitate

w
connection to communication links, a popular standard known as RS-232-C was

developed.
w

4.2 STANDARD I/O INTERFACES
The processor bus is the bus defied by the signals on the processor chip
itself. Devices that require a very high-speed connection to the processor, such as the
om
main memory, may be connected directly to this bus. For electrical reasons, only a few
devices can be connected in this manner. The motherboard usually provides another bus
that can support more devices. The two buses are interconnected by a circuit, which we
c
will call a bridge, that translates the signals and protocols of one bus into those of the
p.
other. Devices connected to the expansion bus appear to the processor as if they were
connected directly to the processors own bus. The only difference is that the bridge
ou
circuit introduces a small delay in data transfers between the processor and those devices.
gr
It is not possible to define a uniform standard for the processor bus. The structure
of this bus is closely tied to the architecture of the processor. It is also dependent on the
ts
electrical characteristics of the processor chip, such as its clock speed. The expansion bus
en
is not subject to these limitations, and therefore it can use a standardized signaling
scheme. A number of standards have been developed. Some have evolved by default,
ud
when a particular design became commercially successful. For example, IBM developed
a bus they called ISA (Industry Standard Architecture) for their personal computer known
at the time as PC AT.
st
Some standards have been developed through industrial cooperative efforts, even
ity
among competing companies driven by their common self-interest in having compatible

products. In some cases, organizations such as the IEEE (Institute of Electrical and
.c
Electronics Engineers), ANSI (American National Standards Institute), or international

bodies such as ISO (International Standards Organization) have blessed these standards
w
and given them an official status.

w
A given computer may use more than one bus standards. A typical Pentium
w
computer has both a PCI bus and an ISA bus, thus providing the user with a wide range
of devices to choose from.

Figure 21 An example of a computer system using different interface standards
Main
om
Memory
c
p.
Processor bus
ou
Processor Bridge
gr
ts PCI bus
en
Additional SCSI Ethernet USB ISA
memory controller interface controller interface
ud
st
ity
Peripheral Component Interconnect (PCI) Bus:-

.c
The PCI bus is a good example of a system bus that grew out of the need for
standardization. It supports the functions found on a processor bus bit in a standardized
w
format that is independent of any particular processor. Devices connected to the PCI bus
w
appear to the processor as if they were connected directly to the processor bus. They are
assigned addresses in the memory address space of the processor.
w
The PCI follows a sequence of bus standards that were used primarily in IBM
PCs. Early PCs used the 8-bit XT bus, whose signals closely mimicked those of Intels

80x86 processors. Later, the 16-bit bus used on the PC At computers became known as
the ISA bus. Its extended 32-bit version is known as the EISA bus. Other buses
developed in the eighties with similar capabilities are the Microchannel used in IBM PCs
and the NuBus used in Macintosh computers.
om
The PCI was developed as a low-cost bus that is truly processor independent. Its
design anticipated a rapidly growing demand for bus bandwidth to support high-speed
c
disks and graphic and video devices, as well as the specialized needs of multiprocessor
p.
systems. As a result, the PCI is still popular as an industry standard almost a decade after
it was first introduced in 1992.
ou
An important feature that the PCI pioneered is a plug-and-play capability for
gr
connecting I/O devices. To connect a new device, the user simply connects the device
interface board to the bus. The software takes care of the rest.
ts
en
Data Transfer:-
In todays computers, most memory transfers involve a burst of data rather than
ud
just one word. The reason is that modern processors include a cache memory. Data are
transferred between the cache and the main memory in burst of several words each. The
st
words involved in such a transfer are stored at successive memory locations. When the
ity
processor (actually the cache controller) specifies an address and requests a read
operation from the main memory, the memory responds by sending a sequence of data
.c
words starting at that address. Similarly, during a write operation, the processor sends a
memory address followed by a sequence of data words, to be written in successive
w
memory locations starting at the address. The PCI is designed primarily to support this
w
mode of operation. A read or write operation involving a single word is simply treated as
a burst of length one.
w
The bus supports three independent address spaces: memory, I/O, and
configuration. The first two are self-explanatory. The I/O address space is intended for

use with processors, such as Pentium, that have a separate I/O address space. However, as
noted , the system designer may choose to use memory-mapped I/O even when a separate
I/O address space is available. In fact, this is the approach recommended by the PCI its
plug-and-play capability. A 4-bit command that accompanies the address identifies which
of the three spaces is being used in a given data transfer operation.
om
The signaling convention on the PCI bus is similar to the one used, we assumed
c
that the master maintains the address information on the bus until data transfer is
p.
completed. But, this is not necessary. The address is needed only long enough for the
slave to be selected. The slave can store the address in its internal buffer. Thus, the
ou
address is needed on the bus for one clock cycle only, freeing the address lines to be used
for sending data in subsequent clock cycles. The result is a significant cost reduction
gr
because the number of wires on a bus is an important cost factor. This approach in used
in the PCI bus. ts
At any given time, one device is the bus master. It has the right to initiate data
en
transfers by issuing read and write commands. A master is called an initiator in PCI
terminology. This is either a processor or a DMA controller. The addressed device that
responds to read and write commands is called a target.
ud
Device Configuration:-
st
When an I/O device is connected to a computer, several actions are needed to

ity
configure both the device and the software that communicates with it.
.c
The PCI simplifies this process by incorporating in each I/O device interface a
small configuration ROM memory that stores information about that device. The
w
configuration ROMs of all devices is accessible in the configuration address space. The
w
PCI initialization software reads these ROMs whenever the system is powered up or
reset. In each case, it determines whether the device is a printer, a keyboard, an Ethernet
w
interface, or a disk controller. It can further learn bout various device options and
characteristics.

Devices are assigned addresses during the initialization process. This means that
during the bus configuration operation, devices cannot be accessed based on their
address, as they have not yet been assigned one. Hence, the configuration address space
uses a different mechanism. Each device has an input signal called Initialization Device
Select, IDSEL#.
om
The PCI bus has gained great popularity in the PC word. It is also used in many
c
other computers, such as SUNs, to benefit from the wide range of I/O devices for which a
p.
PCI interface is available. In the case of some processors, such as the Compaq Alpha, the
PCI-processor bridge circuit is built on the processor chip itself, further simplifying
ou
system design and packaging.
gr
SCSI Bus:-
ts
The acronym SCSI stands for Small Computer System Interface. It refers to a
en
standard bus defined by the American National Standards Institute (ANSI) under the
designation X3.131 . In the original specifications of the standard, devices such as disks
are connected to a computer via a 50-wire cable, which can be up to 25 meters in length
ud
and can transfer data at rates up to 5 megabytes/s.

st
The SCSI bus standard has undergone many revisions, and its data transfer
ity
capability has increased very rapidly, almost doubling every two years. SCSI-2 and
SCSI-3 have been defined, and each has several options. A SCSI bus may have eight data
.c
lines, in which case it is called a narrow bus and transfers data one byte at a time.
Alternatively, a wide SCSI bus has 16 data lines and transfers data 16 bits at a time.
w
There are also several options for the electrical signaling scheme used.
w
Devices connected to the SCSI bus are not part of the address space of the
w
processor in the same way as devices connected to the processor bus. The SCSI bus is
connected to the processor bus through a SCSI controller. This controller uses DMA to
transfer data packets from the main memory to the device, or vice versa. A packet may

contain a block of data, commands from the processor to the device, or status information
about the device.
To illustrate the operation of the SCSI bus, let us consider how it may be used
with a disk drive. Communication with a disk drive differs substantially from
om
communication with the main memory.
A controller connected to a SCSI bus is one of two types an initiator or a target.
c
p.
An initiator has the ability to select a particular target and to send commands specifying
the operations to be performed. Clearly, the controller on the processor side, such as the
ou
SCSI controller, must be able to operate as an initiator. The disk controller operates as a
target. It carries out the commands it receives from the initiator. The initiator establishes
gr
a logical connection with the intended target. Once this connection has been established,
it can be suspended and restored as needed to transfer commands and bursts of data.
ts
While a particular connection is suspended, other device can use the bus to transfer
en
information. This ability to overlap data transfer requests is one of the key features of the
SCSI bus that leads to its high performance.
ud
Data transfers on the SCSI bus are always controlled by the target controller. To
send a command to a target, an initiator requests control of the bus and, after winning
st
arbitration, selects the controller it wants to communicate with and hands control of the
ity
bus over to it. Then the controller starts a data transfer operation to receive a command
from the initiator.
.c
The processor sends a command to the SCSI controller, which causes the
w
following sequence of event to take place:

w
1. The SCSI controller, acting as an initiator, contends for control of the bus.
w
2. When the initiator wins the arbitration process, it selects the target controller and
hands over control of the bus to it.

3. The target starts an output operation (from initiator to target); in response to this,
the initiator sends a command specifying the required read operation.
4. The target, realizing that it first needs to perform a disk seek operation, sends a
message to the initiator indicating that it will temporarily suspend the connection
between them. Then it releases the bus.
om
5. The target controller sends a command to the disk drive to move the read head to
the first sector involved in the requested read operation. Then, it reads the data
c
stored in that sector and stores them in a data buffer. When it is ready to begin
p.
transferring data to the initiator, the target requests control of the bus. After it
wins arbitration, it reselects the initiator controller, thus restoring the suspended
ou
connection.
6. The target transfers the contents of the data buffer to the initiator and then
gr
suspends the connection again. Data are transferred either 8 or 16 bits in parallel,
depending on the width of the bus.
ts
7. The target controller sends a command to the disk drive to perform another seek
en
operation. Then, it transfers the contents of the second disk sector to the initiator
as before. At the end of this transfers, the logical connection between the two
controllers is terminated.
ud
8. As the initiator controller receives the data, it stores them into the main memory
using the DMA approach.
st
9. The SCSI controller sends as interrupt to the processor to inform it that the
ity
requested operation has been completed

.c
This scenario show that the messages exchanged over the SCSI bus are at a higher
level than those exchanged over the processor bus. In this context, a higher level means
w
that the messages refer to operations that may require several steps to complete,
w
depending on the device. Neither the processor nor the SCSI controller need be aware of
the details of operation of the particular device involved in a data transfer. In the
w
preceding example, the processor need not be involved in the disk seek operation.

UNIVERSAL SERIAL BUS (USB):-
The synergy between computers and communication is at the heart of todays

information technology revolution. A modern computer system is likely to involve a wide
variety of devices such as keyboards, microphones, cameras, speakers, and display
om
devices. Most computers also have a wired or wireless connection to the Internet. A key
requirement is such an environment is the availability of a simple, low-cost mechanism to
c
connect these devices to the computer, and an important recent development in this
p.
regard is the introduction of the Universal Serial Bus (USB). This is an industry standard
developed through a collaborative effort of several computer and communication
ou
companies, including Compaq, Hewlett-Packard, Intel, Lucent, Microsoft, Nortel
Networks, and Philips.
gr
The USB supports two speeds of operation, called low-speed (1.5 megabits/s)
ts
and full-speed (12 megabits/s). The most recent revision of the bus specification (USB
en
2.0) introduced a third speed of operation, called high-speed (480 megabits/s). The USB
is quickly gaining acceptance in the market place, and with the addition of the high-speed
capability it may well become the interconnection method of choice for most computer
ud
devices.
st
The USB has been designed to meet several key objectives:

ity
Provides a simple, low-cost and easy to use interconnection system that

overcomes the difficulties due to the limited number of I/O ports available on a
.c
computer.
Accommodate a wide range of data transfer characteristics for I/O devices,
w
including telephone and Internet connections.

w
Enhance user convenience through a plug-and-play mode of operation

w

Port Limitation:-
The parallel and serial ports described in previous section provide a general-
purpose point of connection through which a variety of low-to medium-speed devices can
be connected to a computer. For practical reasons, only a few such ports are provided in a
om
typical computer.
c
Device Characteristics:-
p.
The kinds of devices that may be connected to a computer cover a wide range of
ou
functionality. The speed, volume, and timing constraints associated with data transfers to
and from such devices vary significantly.
gr
A variety of simple devices that may be attached to a computer generate data of a
similar nature low speed and asynchronous. Computer mice and the controls and
ts
manipulators used with video games are good examples.
en
Plug-and-Play:-
ud
As computers become part of everyday life, their existence should become

increasingly transparent. For example, when operating a home theater system, which
st
includes at least one computer, the user should not find it necessary to turn the computer
ity
off or to restart the system to connect or disconnect a device.

The plug-and-play feature means that a new device, such as an additional
.c
speaker, can be connected at any time while the system is operating. The system should
detect the existence of this new device automatically, identify the appropriate device-
w
driver software and any other facilities needed to service that device, and establish the
w
appropriate addresses and logical connections to enable them to communicate. The plug-
and-play requirement has many implications at all levels in the system, from the
w
hardware to the operating system and the applications software. One of the primary
objectives of the design of the USB has been to provide a plug-and-play capability.

USB Architecture:-
The discussion above points to the need for an interconnection system that
combines low cost, flexibility, and high data-transfer bandwidth. Also, I/O devices may
be located at some distance from the computer to which they are connected. The
requirement for high bandwidth would normally suggest a wide bus that carries 8, 16, or
om
more bits in parallel. However, a large number of wires increases cost and complexity
and is inconvenient to the user. Also, it is difficult to design a wide bus that carries data
c
for a long distance because of the data skew problem discussed. The amount of skew
p.
increases with distance and limits the data that can be used.
A serial transmission format has been chosen for the USB because a serial bus
ou
satisfies the low-cost and flexibility requirements. Clock and data information are
encoded together and transmitted as a single signal. Hence, there are no limitations on
gr
clock frequency or distance arising from data skew. Therefore, it is possible to provide a
high data transfer bandwidth by using a high clock frequency. As pointed out earlier, the
ts
USB offers three bit rates, ranging from 1.5 to 480 megabits/s, to suit the needs of
en
different I/O devices.
Figure 23 Universal Serial Bus tree structure.

ud
st
Host computer
Root
ity
hub
.c
Hub Hub
w
w
I/O I/O I/O I/O

device device device device
w
Hub
I/O I/O
device device

To accommodate a large number of devices that can be added or removed at any

time, the USB has the tree structure shown in figure 23. Each node of the tree has a
device called a hub, which acts as an intermediate control point between the host and the
I/O devices. At the root of the tree, a root hub connects the entire tree to the host
om
computer. The leaves of the tree are the I/O devices being served (for example, keyboard,
Internet connection, speaker, or digital TV), which are called functions in USB
c
terminology. For consistency with the rest of the discussion in the book, we will refer to
p.
these devices as I/O devices.
ou
The tree structure enables many devices to be connected while using only simple
point-to-point serial links. Each hub has a number of ports where devices may be
gr
connected, including other hubs. In normal operation, a hub copies a message that it
receives from its upstream connection to all its downstream ports. As a result, a message
ts
sent by the host computer is broadcast to all I/O devices, but only the addressed device
en
will respond to that message. In this respect, the USB functions in the same way as the
bus in figure 4.1. However, unlike the bus in figure 4.1, a message from an I/O device is
sent only upstream towards the root of the tree and is not seen by other devices. Hence,
ud
the USB enables the host to communicate with the I/O devices, but it does not enable
these devices to communicate with each other.
st
ity
Note how the tree structure helps meet the USBs design objectives. The tree
makes it possible to connect a large number of devices to a computer through a few ports
.c
(the root hub). At the same time, each I/O device is connected through a serial point-to-
point connection. This is an important consideration in facilitating the plug-and-play
w
feature, as we will see shortly.

w
The USB operates strictly on the basis of polling. A device may send a message
only in response to a poll message from the host. Hence, upstream messages do not
w
encounter conflicts or interfere with each other, as no two devices can send messages at
the same time. This restriction allows hubs to be simple, low-cost devices.

The mode of operation described above is observed for all devices operating at
either low speed or full speed. However, one exception has been necessitated by the
introduction of high-speed operation in USB version 2.0. Consider the situation in figure
24. Hub A is connected to the root hub by a high-speed link. This hub serves one high-
speed device, C, and one low-speed device, D. Normally, a messages to device D would
om
be sent at low speed from the root hub. At 1.5 megabits/s, even a short message takes
several tens of microsends. For the duration of this message, no other data transfers can
c
take place, thus reducing the effectiveness of the high-speed links and introducing
p.
unacceptable delays for high-speed devices. To mitigate this problem, the USB protocol
requires that a message transmitted on a high-speed link is always transmitted at high
ou
speed, even when the ultimate receiver is a low-speed device. Hence, a message at low
speed to device D. The latter transfer will take a long time, during which high-speed
gr
traffic to other nodes is allowed to continue. For example, the root hub may exchange
several message with device C while the low-speed message is being sent from hub A to
ts
device D. During this period, the bus is said to be split between high-speed and low-
en
speed traffic. The message to device D is preceded and followed by special commands to
hub A to start and end the split-traffic mode of operation, respectively.
ud
st
ity
.c
w
w
w

UNIT - 5
Memory System: Basic Concepts, Semiconductor RAM Memories, Read
Only Memories, Speed, Size, and Cost, Cache Memories Mapping
om
Functions, Replacement Algorithms, Performance Considerations, Virtual
Memories, Secondary Storage
c
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
w

UNIT - 5
MEMORY SYSTEM
5.1 BASIC CONCEPTS:
om
The maximum size of the Main Memory (MM) that can be used in any computer
is determined by its addressing scheme. For example, a 16-bit computer that generates
c
16-bit addresses is capable of addressing upto 216 =64K memory locations. If a machine
p.
generates 32-bit addresses, it can access upto 232 = 4G memory locations. This number
represents the size of address space of the computer.
ou
If the smallest addressable unit of information is a memory word, the machine is
gr
called word-addressable. If individual memory bytes are assigned distinct addresses, the
computer is called byte-addressable. Most of the commercial machines are byte-
ts
addressable. For example in a byte-addressable 32-bit computer, each memory word
en
contains 4 bytes. A possible word-address assignment would be:
Word Address Byte Address

ud
0 0 1 2 3
4 4 5 6 7
st
8 8 9 10 11
ity
. ..
. ..
.c
. ..
With the above structure a READ or WRITE may involve an entire memory
w
word or it may involve only a byte. In the case of byte read, other bytes can also be read
w
but ignored by the CPU. However, during a write cycle, the control circuitry of the MM
must ensure that only the specified byte is altered. In this case, the higher-order 30 bits
w
can specify the word and the lower-order 2 bits can specify the byte within the word.

CPU-Main Memory Connection A block schematic: -

From the system standpoint, the Main Memory (MM) unit can be viewed as a
block box. Data transfer between CPU and MM takes place through the use of two
CPU registers, usually called MAR (Memory Address Register) and MDR (Memory Data
Register). If MAR is K bits long and MDR is n bits long, then the MM unit may contain
om
upto 2k addressable locations and each location will be n bits wide, while the word
length is equal to n bits. During a memory cycle, n bits of data may be transferred
c
between the MM and CPU. This transfer takes place over the processor bus, which has k
p.
address lines (address bus), n data lines (data bus) and control lines like Read, Write,
Memory Function completed (MFC), Bytes specifiers etc (control bus). For a read
ou
operation, the CPU loads the address into MAR, set READ to 1 and sets other control
signals if required. The data from the MM is loaded into MDR and MFC is set to 1. For a
gr
write operation, MAR, MDR are suitably loaded by the CPU, write is set to 1 and other
control signals are set suitably. The MM control circuitry loads the data into appropriate
ts
locations and sets MFC to 1. This organization is shown in the following block schematic
en
Address Bus (k bits)
ud
MAR Main Memory upto 2k

addressable locations
st
Word length = n bits

Data bus (n bits)
ity
MDR
.c
CPU Control Bus

w
w
(Read, Write, MFC, Byte

Specifier etc)
w

Some Basic Concepts

Memory Access Times: -
It is a useful measure of the speed of the memory unit. It is the time that elapses
between the initiation of an operation and the completion of that operation (for example,
the time between READ and MFC).
om
Memory Cycle Time :-
c
It is an important measure of the memory system. It is the minimum time delay
p.
required between the initiations of two successive memory operations (for example, the
time between two successive READ operations). The cycle time is usually slightly longer
ou
than the access time.
gr
5.2 RANDOM ACCESS MEMORY (RAM):
A memory unit is called a Random Access Memory if any location can be
ts
accessed for a READ or WRITE operation in some fixed amount of time that is
en
independent of the locations address. Main memory units are of this type. This
distinguishes them from serial or partly serial access storage devices such as magnetic
ud
tapes and disks which are used as the secondary storage device.
Cache Memory:-
st
The CPU of a computer can usually process instructions and data faster than they
ity
can be fetched from compatibly priced main memory unit. Thus the memory cycle time
become the bottleneck in the system. One way to reduce the memory access time is to use
.c
cache memory. This is a small and fast memory that is inserted between the larger,
slower main memory and the CPU. This holds the currently active segments of a program
w
and its data. Because of the locality of address references, the CPU can, most of the time,
w
find the relevant information in the cache memory itself (cache hit) and infrequently
needs access to the main memory (cache miss) with suitable size of the cache memory,
w
cache hit rates of over 90% are possible leading to a cost-effective increase in the
performance of the system.

Memory Interleaving: -
This technique divides the memory system into a number of memory modules
and arranges addressing so that successive words in the address space are placed in
different modules. When requests for memory access involve consecutive addresses, the
access will be to different modules. Since parallel access to these modules is possible, the
om
average rate of fetching words from the Main Memory can be increased.
c
Virtual Memory: -
p.
In a virtual memory System, the address generated by the CPU is referred to as a
virtual or logical address. The corresponding physical address can be different and the
ou
required mapping is implemented by a special memory control unit, often called the
memory management unit. The mapping function itself may be changed during program
gr
execution according to system requirements.
ts
Because of the distinction made between the logical (virtual) address space and
en
the physical address space; while the former can be as large as the addressing capability
of the CPU, the actual physical memory can be much smaller. Only the active portion of
the virtual address space is mapped onto the physical memory and the rest of the virtual
ud
address space is mapped onto the bulk storage device used. If the addressed information
is in the Main Memory (MM), it is accessed and execution proceeds. Otherwise, an
st
exception is generated, in response to which the memory management unit transfers a

ity
contigious block of words containing the desired word from the bulk storage unit to the
MM, displacing some block that is currently inactive. If the memory is managed in such a
.c
way that, such transfers are required relatively infrequency (ie the CPU will generally
find the required information in the MM), the virtual memory system can provide a
w
reasonably good performance and succeed in creating an illusion of a large memory with
w
a small, in expensive MM.

w

5.3 INTERNAL ORGANIZATION OF SEMICONDUCTOR

MEMORY CHIPS:
Memory chips are usually organized in the form of an array of cells, in which
om
each cell is capable of storing one bit of information. A row of cells constitutes a memory
word, and the cells of a row are connected to a common line referred to as the word line,
and this line is driven by the address decoder on the chip. The cells in each column are
c
connected to a sense/write circuit by two lines known as bit lines. The sense/write circuits
p.
are connected to the data input/output lines of the chip. During a READ operation, the
Sense/Write circuits sense, or read, the information stored in the cells selected by a word
ou
line and transmit this information to the output lines. During a write operation, they
receive input information and store it in the cells of the selected word.
gr
The following figure shows such an organization of a memory chip consisting of
ts
16 words of 8 bits each, which is usually referred to as a 16 x 8 organization.
en
The data input and the data output of each Sense/Write circuit are connected to a
ud
single bi-directional data line in order to reduce the number of pins required. One control
line, the R/W (Read/Write) input is used a specify the required operation and another
control line, the CS (Chip Select) input is used to select a given chip in a multichip
st
memory system. This circuit requires 14 external connections, and allowing 2 pins for
ity
power supply and ground connections, can be manufactured in the form of a 16-pin chip.
It can store 16 x 8 = 128 bits.
.c
Another type of organization for 1k x 1 format is shown below:

w
w
w

5 Bit Row Address W0
W1
5-Bit 32X32
om
Decoder memory cell
array
c
10
B
p.
W31
I
T
ou
A Two 32 to 1 Sence/Write
D Multiplexers
D Circuitr
(input & 1 output)
gr
R
E
S ts
S
en
The 10-bit address is divided into two groups of 5 bits each to form the row and column
addresses for the cell array. A row address selects a row of 32 cells, all of which are
accessed in parallel. One of these, selected by the column address, is connected to the
ud
external data lines by the input and output multiplexers. This structure can store 1024
bits, can be implemented in a 16-pin chip.
st
ity
A Typical Memory Cell

.c
Semiconductor memories may be divided into bipolar and MOS types. They may
be compared as follows:
w
w
Characteristic Bipolar MOS

w
Power Dissipation More Less

Bit Density Less More
Impedance Lower Higher

Speed More Less
Bipolar Memory Cell
A typical bipolar storage cell is shown below:
c om
p.
ou
gr
Two transistor inverters connected to implement a basic flip-flop. The cell is connected to
ts
one word line and two bits lines as shown. Normally, the bit lines are kept at about 1.6V,
en
and the word line is kept at a slightly higher voltage of about 2.5V. Under these
conditions, the two diodes D1 and D2 are reverse biased. Thus, because no current flows
through the diodes, the cell is isolated from the bit lines.
ud
Read Operation:-
st
Let us assume the Q1 on and Q2 off represents a 1 to read the contents of a given
ity
cell, the voltage on the corresponding word line is reduced from 2.5 V to approximately
0.3 V. This causes one of the diodes D1 or D2 to become forward-biased, depending on
.c
whether the transistor Q1 or Q2 is conducting. As a result, current flows from bit line b
when the cell is in the 1 state and from bit line b when the cell is in the 0 state. The
w
Sense/Write circuit at the end of each pair of bit lines monitors the current on lines b and
w
b and sets the output bit line accordingly.

w
Write Operation: -
While a given row of bits is selected, that is, while the voltage on the
corresponding word line is 0.3V, the cells can be individually forced to either the 1 state

by applying a positive voltage of about 3V to line b or to the 0 state by driving line b.

This function is performed by the Sense/Write circuit.
MOS Memory Cell: -
MOS technology is used extensively in Main Memory Units. As in the case of
om
bipolar memories, many MOS cell configurations are possible. The simplest of these is a
flip-flop circuit. Two transistors T1 and T2 are connected to implement a flip-flop.
c
Active pull-up to VCC is provided through T3 and T4. Transistors T5 and T6 act as
p.
switches that can be opened or closed under control of the word line. For a read
operation, when the cell is selected, T5 or T6 is closed and the corresponding flow of
ou
current through b or b is sensed by the sense/write circuits to set the output bit line
accordingly. For a write operation, the bit is selected and a positive voltage is applied on
gr
the appropriate bit line, to store a 0 or 1. This configuration is shown below:
ts
en
ud
st
ity
Static Memories Vs Dynamic Memories:-

.c
Bipolar as well as MOS memory cells using a flip-flop like structure to store
w
information can maintain the information as long as current flow to the cell is maintained.
w
Such memories are called static memories. In contracts, Dynamic memories require not
only the maintaining of a power supply, but also a periodic refresh to maintain the
w
information stored in them. Dynamic memories can have very high bit densities and very
lower power consumption relative to static memories and are thus generally used to
realize the main memory unit.

Dynamic Memories:-
The basic idea of dynamic memory is that information is stored in the form of a
charge on the capacitor. An example of a dynamic memory cell is shown below:
om
When the transistor T is turned on and an appropriate voltage is applied to the bit
c
line, information is stored in the cell, in the form of a known amount of charge stored on
p.
the capacitor. After the transistor is turned off, the capacitor begins to discharge. This is
caused by the capacitors own leakage resistance and the very small amount of current
ou
that still flows through the transistor. Hence the data is read correctly only if is read
before the charge on the capacitor drops below some threshold value. During a Read
gr
ts
en
ud
operation, the bit line is placed in a high-impendance state, the transistor is turned on and
a sense circuit connected to the bit line is used to determine whether the charge on the
st
capacitor is above or below the threshold value. During such a Read, the charge on the
ity
capacitor is restored to its original value and thus the cell is refreshed with every read
operation.
.c
w
w
w

Typical Organization of a Dynamic Memory Chip:-

RAS
Row 4096*(512*8
Row Decoder )
Address Cell array
L t h
c om
Sense/Write
circuits
p.
A7-0 CS
ou
Column Column DI/D0
Address Decoder
gr
R/W
CAS ts
en
A typical organization of a 64k x 1 dynamic memory chip is shown below:
The cells are organized in the form of a square array such that the high-and
lower-order 8 bits of the 16-bit address constitute the row and column addresses of a cell,
ud
respectively. In order to reduce the number of pins needed for external connections, the
row and column address are multiplexed on 8 pins. To access a cell, the row address is
st
applied first. It is loaded into the row address latch in response to a single pulse on the
ity
Row Address Strobe (RAS) input. This selects a row of cells. Now, the column address is
applied to the address pins and is loaded into the column address latch under the control
.c
of the Column Address Strobe (CAS) input and this address selects the appropriate
sense/write circuit. If the R/W signal indicates a Read operation, the output of the
w
selected circuit is transferred to the data output. Do. For a write operation, the data on the
w
DI line is used to overwrite the cell selected.

w
It is important to note that the application of a row address causes all the cells on
the corresponding row to be read and refreshed during both Read and Write operations.
To ensure that the contents of a dynamic memory are maintained, each row of cells must

be addressed periodically, typically once every two milliseconds. A Refresh circuit

performs this function. Some dynamic memory chips incorportate a refresh facility the
chips themselves and hence they appear as static memories to the user! such chips are
often referred to as Pseudostatic.
om
Another feature available on many dynamic memory chips is that once the row
address is loaded, successive locations can be accessed by loading only column
c
addresses. Such block transfers can be carried out typically at a rate that is double that for
p.
transfers involving random addresses. Such a feature is useful when memory access
follow a regular pattern, for example, in a graphics terminal.
ou
Because of their high density and low cost, dynamic memories are widely used in
gr
the main memory units of computers. Commercially available chips range in size from 1k
to 4M bits or more, and are available in various organizations like 64k x 1, 16k x 4, 1MB
ts
x 1 etc.
en
DESIGN CONSIDERATION FOR MEMORY SYSTEMS:-
ud
The choice of a RAM chip for a given application depends on several factors like
speed, power dissipation, size of the chip, availability of block transfer feature etc.
st
ity
Bipolar memories are generally used when very fast operation is the primary
requirement. High power dissipation in bipolar circuits makes it difficult to realize high
.c
bit densities.
w
Dynamic MOS memory is the predominant technology used in the main

w
memories of computer, because their high bit-density makes it possible to implement

large memories economically.
w

Static MOS memory chips have higher densities and slightly longer access times
compared to bipolar chips. They have lower densities than dynamic memories but are
easier to use because they do not require refreshing.
Design using static Memory Chips: -
om
Consider the design of a memory system of 64k x 16 using 16k x 1 static
c
memory chips. We can use a column of 4 chips to implement one bit position. Sixteen
p.
such sets provide the required 64k x 16 memories. Each chip has a control input called
chip select, which should be set to 1 to enable the chip to participate in a read or write
ou
operation. When this signal is 0, the chip is electrically isolated from the rest of the
system. The high-order 2 bits of the 16-bit address bus are decoded to obtain the four
gr
chip select control signals, and the remaining 14address bits are used to access specific

bit locations inside each chip of the selected row. The R
ts /W input of all chips are fed
together to provide a common Read/ Write control. This organization is shown in the
en
following figure.
ud
st
ity
.c
w
w
w

21-bit
addresses
A0
om
A1
..
c
p.
ou
A19
A20
gr
ts
en
ud
st
ity
.c
21-bit
decoder
w
w
512 K x 8
w
memory chip
D31-24 D23-16 D15-8 D7-0

Design using Dynamic Memory Chips: -

The design of a memory system using dynamic memory chips is similar to the
design using static memory chips, but with the following important differences in the
control circry.
om
The row and column parts of the address of each chip have to be multiplexed;
A refresh circuit is needed; and
c
The timing of various steps of a memory cycle must be carefully controlled.
p.
A memory system of 256k x 16 designed using 64k x 1 DRAM chips, is shown below;
ou
gr
ts
en
ud
st
ity
.c
The memory unit is assumed to be connected to an asynchronous memory bus that has
w
18 address lines (ADRS 17-0), 16 data lines (DATA15-0), two handshake signals (Memory
w
request and MFC), and a Read/ Write line to specify the operation (read to Write).
w
The memory chips are organized into 4 rows, with each row having 16 chips.
Thus each column of the 16 columns implements one bit position. The higher order 12
bits of the address are decoded to get four chip select control signals which are used to

select one of the four rows. The remaining 16 bits, after multiplexing, are used to access
specific bit locations inside each chip of the selected row. The R/W inputs of all chips are
tied together to provide a common Read/Write control. The DI and DO lines, together,
provide D15 D0 i.e. the data bus DATA15-0.
om
The operation of the control circuit can be described by considering a
normal memory read cycle. The cycle begins when the CPU activates the address, the
Read/ Write and the Memory Request lines. When the memory request line becomes
c
p.
active, the Access control block sets the start signal to 1. The timing control block
responds by setting Busy lines to 1, in order to prevent the access control block from
ou
accepting new requests until the current cycle ends. The timing control block then loads
the row and column addresses into the memory chips by activating the RAS and CAS
gr
lines. During this time, it uses the Row/ Column line to select first the row address,
ADRS15-8, followed by the column address, ADRS(7-0).
ts
en
Refresh Operation:-
ud
The Refresh control block periodically generates Refresh, requests, causing the
access control block to start a memory cycle in the normal way. This block allows the
refresh operation by activating the Refresh Grant line. The access control block arbitrates
st
between Memory Access requests and Refresh requests, with priority to Refresh requests
ity
in the case of a tie to ensure the integrity of the stored data.

As soon as the Refresh control block receives the Refresh Grant signal, it
.c
activates the Refresh line. This causes the address multiplexer to select the Refresh
counter as the source and its contents are thus loaded into the row address latches of all
w
memory chips when the RAS signal is activated. During this time the R/ W line may be
w
low, causing an inadvertent write operation. One way to prevent this is to use the Refresh
w
line to control the decoder block to deactivate all the chip select lines. The rest of the
refresh cycle is the same as in a normal cycle. At the end, the Refresh control block
increments the refresh counter in preparation for the next Refresh cycle.

Even though the row address has 8 bits, the Refresh counter need only be 7 bits
wide because of the cell organization inside the memory chips. In a 64k x 1 memory chip,
the 256x256 cell array actually consists of two 128x256 arrays. The low order 7 bits of
the row address select a row from both arrays and thus the row from both arrays is
refreshed!
om
Ideally, the refresh operation should be transparent to the CPU. However, the
c
response of the memory to a request from the CPU or from a DMA device may be
p.
delayed if a Refresh operation is in progress. Further, in the case of a tie, Refresh
operation is given priority. Thus the normal access may be delayed. This delay will be
ou
more if all memory rows are refreshed before the memory is returned to normal use. A
more common scheme, however, interleaves Refresh operations on successive rows with
gr
accesses from the memory bus. In either case, Refresh operations generally use less than
5% of the available memory cycles, so the time penalty caused by refreshing is very
ts
small.
en
The variability in the access times resulting from refresh can be easily
accommodated in an asynchronous bus scheme. With synchronous buses, it may be
ud
necessary for the Refresh circuit to request bus cycles as a DMA device would!
st
Semi Conductor Rom Memories: -

ity
Semiconductor read-only memory (ROM) units are well suited as the control
store components in micro programmed processors and also as the parts of the main
.c
memory that contain fixed programs or data.

The following figure shows a possible configuration for a bipolar ROM cell.
w
w
w
The word line is normally held at a low voltage. If a word is to be selected, the voltage of
the corresponding word line is momentarily raised, which causes all transistors whose

emitters are connected to their corresponding bit lines to be turned on. The current that
flows from the voltage supply to the bit line can be detected by a sense circuit. The bit
positions in which current is detected are read as 1s, and the remaining bits are read as Os.
Therefore, the contents of a given word are determined by the pattern of emitter to bit-
line connections similar configurations are possible in MOS technology.
om
Data are written into a ROM at the time of manufacture programmable ROM
c
(PROM) devices allow the data to be loaded by the user. Programmability is achieved by
p.
connecting a fuse between the emitter and the bit line. Thus, prior to programming, the
memory contains all 1s. The user can inserts Os at the required locations by burning out
ou
the fuses at these locations using high-current pulses. This process is irreversible.
gr
ROMs are attractive when high production volumes are involved. For smaller
numbers, PROMs provide a faster and considerably less expensive approach.
ts
en
Chips which allow the stored data to be erased and new data to be loaded. Such a
chip is an erasable, programmable ROM, usually called an EPROM. It provides
considerable flexibility during the development phase. An EPROM cell bears
ud
considerable resemblance to the dynamic memory cell. As in the case of dynamic

memory, information is stored in the form of a charge on a capacitor. The main
st
difference is that the capacitor in an EPROM cell is very well insulated. Its rate of
ity
discharge is so low that it retains the stored information for very long periods. To write
information, allowing charge to be stored on the capacitor.
.c
The contents of EPROM cells can be erased by increasing the discharge rate of
w
the storage capacitor by several orders of magnitude. This can be accomplished by

w
allowing ultraviolet light into the chip through a window provided for that purpose, or by
the application of a high voltage similar to that used in a write operation. If ultraviolet
w
light is used, all cells in the chip are erased at the same time. When electrical erasure is
used, however, the process can be made selective. An electrically erasable EPROM, often
referred to as EEPROM. However, the circuit must now include high voltage generation.

Some E2PROM chips incorporate the circuitry for generating these voltages o the chip
itself. Depending on the requirements, suitable device can be selected.
Classification of memory devices
Memory devise
c om
Read/Write Read only
p.
ou
Static Dynamic ROM PROM Erasable PROM
gr
Bi-polar MOS ts UV Erasable PROM E2PROM
en
5.5 CACHE MEMORY:
Analysis of a large number of typical programs has shown that most of their
ud
execution time is spent on a few main row lines in which a number of instructions are
executed repeatedly. These instructions may constitute a simple loop, nested loops or few
procedure that repeatedly call each other. The main observation is that many instructions
st
in a few localized are as of the program are repeatedly executed and that the remainder of
ity
the program is accessed relatively infrequently. This phenomenan is referred to as

locality of reference.
.c
If the active segments of a program can be placed in a fast memory, then the total
execution time can be significantly reduced, such a memory is referred as a cache
w
memory which is in served between the CPU and the main memory as shown in fig.1
w
Fig.1 cache memory between main memory & cpu.

w
CPU CACHE Main memory

Two Level memory Hierarchy: We will adopt the terms Primary level for the
smaller, faster memory and the secondary level for larger, slower memory, we will also
allow cache to be a primary level with slower semiconductor memory as the
corresponding secondary level. At a different point in the hierarchy, the same S.C
memory could be the primary level with disk as the secondary level.
om
Primary and Secondary addresses:-
c
p.
A two level hierarchy and its addressing are illustrated in fig.2. A system address
is applied to the memory management unit (MMU) that handles the mapping function for
ou
the particular pair in the hierarchy. If the MMU finds the address in the Primary level, it
provides Primary address, which selects the item from the Primary memory. This
gr
translation must be fast, because every time memory is accessed, the system address must
be translated. The translation may fail to produce a Primary address because the
ts
requested items is not found, so that information can be retrieved from the secondary
en
level and transferred to the Primary level.
ud
st
ity
.c
w
Hits and Misses:- Successful translation of reference into Primary address is called a
w
hit, and failure is a miss. The hit ratio is (1-miss ratio). If tp is the Primary memory access
w
time and ts is the secondary access time, the average access time for the two level
hierarchy is
ta = h tp + (1-h)ts

The Cache:- The Mapping Function
Fig. Shows a schematic view of the cache mapping function. The mapping
function is responsible for all cache operations. It is implemented in hardware, because of
om
the required high speed operation. The mapping function determines the following.
c
Placement strategies Where to place an incoming block in the cache
p.
Replacement strategies Which block to replace when a cache miss occurs
How to handle Read and Writes up as cache hit and mises
ou
CPU Main
Cache
gr
Word Block Memor
y
ts
en
Mapping function
ud
st
ity
Three different types of mapping functions are in common use

1. Associative mapping cache
.c
2. Direct mapping cache

3. Block-set-associative mapping cache
w
w
1. Associative mapped caches:- In this any block from main memory can be placed any
where in the cache. After being placed in the cache, a given block is identified uniquely
w
by its main memory block number, referred to as the tag, which is stored inside a separate
tag memory in the cache.

Regardless of the kind of cache, a given block in the cache may or may not
contain valid information. For example, when the system has just been powered up add
before the cache has had any blocks loaded into it, all the information there is invalid.
The cache maintains a valid bit for each block to keep track of whether the information in
the corresponding block is valid.
om
Fig4. shows the various memory structures in an associative cache, The cache
c
itself contain 256, 8byte blocks, a 256 x 13 bit tag memory for holding the tags of the
p.
blocks currently stored in the cache, and a 256 x 1 bit memory for storing the valid bits,
Main memory contains 8192, 8 byte blocks. The figure indicates that main memory
ou
address references are partition into two fields, a 3 bit word field describing the location
of the desired word in the cache line, and a 13 bit tag field describing the main memory
gr
block number desired. The 3 bit word field becomes essentially a cache address
specifying where to find the word if indeed it is in the cache.
ts
en
The remaining 13 bits must be compared against every 13 bit tag in the tag
memory to see if the desired word is present.
ud
st
ity
.c
w
w
w
In the fig, above, main memory block 2 has been stored in the 256 cache block
and so the 256th tag entry is 2 mm block 119 has been stored in the second cache block

corresponding entry in tag memory is 119 mm block 421 has been stored in cache block 0
and tag memory location 0 has been set to 421. Three valid bits have also been set,
indicating valid information in these locations.
The associative cache makes the most flexible and complete use of its capacity,
om
storing blocks wherever it needs to, but there is a penalty to be paid for this flexibility the
tag memory must be searched in for each memory reference.
c
p.
Fig.5 shows the mechanism of operation of the associative cache. The process
begins with the main memory address being placed in the argument register of the
ou
(associative) tag memory (1) if there is a match (hit), (2) and if the ratio bit for the block
is set (3), then the block is gated out of the cache (4), and the 3 bit offset field is used to
gr
select the byte corresponding to the block offset field of the main memory address (5)
That byte is forwarded to the CPU, (6)
ts
en
ud
st
ity
.c
w
2. Direct mapped caches:- In this a given main memory block can be placed in one and
w
only one place in the cache. Fig6. Shows an example of a direct mapped cache. For
simplicity, the example again uses a 256 block x 8 byte cache and a 16 bit main memory
w
address. The main memory in the fig. has 256 rows x 32 columns, still fielding 256 x 32
= 8192 = 213 total blocks as before. Notice that the main memory address is partitioned
into 3 fields. The word field still specifies the word in the block. The group field specifies

which of the 256 cache locations the block will be in, if it is indeed in the cache. The tag
field specifies which of the 32 blocks from main memory is actually present in the cache.
Now the cache address is composed of the group field, which specifies the address of the
block location in the cache and the word field, which specifies the address of the word in
the block. There is also valid bit specifying whether the information in the selected block
om
in valid.
c
p.
ou
gr
ts
en
The fig.6 shows block 7680, from group 0 of MM placed in block location 0 of
ud
the cache and the corresponding tag set to 30. Similarly MM block 259 is in MM group
2, column 1, it is placed in block location 2 of the cache and the corresponding tag
st
memory entry is 1.
ity
The tasks required of the direct mapped cache in servicing a memory request
are shown in fig7.
.c
w
The fig. shows the group field of the memory address being decoded 1) and used
to select the tag of the one cache block location in which the block must be stored if it is
w
the cache. If the valid bit for that block location is gated (2), then that tag is gated out, (3)
w
and compared with the tag of the incoming memory address (4), A cache hit gates the
cache block out (5) and the word field selects the specified word from the block (6), only
one tag needs to be compared, resulting in considerably less hardware than in the
associative memory case.

c om
p.
ou
gr
ts
The direct mapped cache has the advantage of simplicity, but the obvious disadvantage
en
that only a single block from a given group can be present in the cache at any given time.
ud
3. Block-set-Associative cache:-
Block-set-Associative caches share properties of both of the previous mapping

st
functions. It is similar to the direct-mapped cache, but now more than one block from a
ity
given group in main memory can occupy the same group in the cache. Assume the same
main memory and block structure as before, but with the cache being twice as large, so
.c
that a set of two main memory blocks from the same group can occupy a given cache
group.
w
w
Fig 8 shows a 2 way set associative cache that is similar to the direct mapped
w
cache in the previous example, but with twice as many blocks in the cache, arranged so
that a set of any two blocks from each main memory group can be stored in the cache.
MM is still partitioned into an 8 bit set field and a 5 bit tag field, but now there are two

possible places in which a given block can reside and both must be searched
associatively.
The cache group address is the same as that of the direct matched cache, an 8 bit
block location and a 3 bit word location. Fig 8 shows that the cache entries corresponding
om
to the second group contains blocks 513 and 2304. the group field now called the set
field, is again decoded and directs the search to the correct group and new only the tags in
c
the selected group must be searched. So instead of 256 compares, the cache only needs to
p.
do 2.
ou
For simplicity, the valid bits are not shown, but they must be present. The cache
hard ware would be similar so that shown in fig 7. but there would be two simultaneous
gr
comparisons of the two blocks in the set.
ts
en
ud
st
ity
.c
w
w
w

Solved Problems:-
1. A block set associative cache consists of a total of 64 blocks divided into 4 block
sets. The MM contains 4096 blocks each containing 128 words.
a) How many bits are there in MM address?
om
b) How many bits are there in each of the TAG, SET & word fields
c
Solution:- Number of sets = 64/4 = 16
p.
Set bits = 4(24 = 16)
Number of words = 128
ou
Word bits = 7 bits (27 = 128)
MM capacity : 4096 x 128 (212 x 27 = 219)
gr
a) Number of bits in memory address = 19 bits
b) ts
8 4 7
en
TAG SET WORD
TAG bits = 19 (7+4) = 8 bits.
ud
2. A computer system has a MM capacity of a total of 1M 16 bits words. It also has

a 4K words cache organized in the block set associative manner, with 4 blocks per
st
set & 64 words per block. Calculate the number of bits in each of the TAG, SET
ity
& WORD fields of MM address format.

.c
Solution: Capacity: 1M (220 = 1M)

Number of words per block = 64
w
Number of blocks in cache = 4k/64 = 64

w
Number of sets = 64/4 = 16

Set bits = 4 (24 = 16)
w
Word bits = 6 bits (26 = 64)

Tag bits = 20-(6+4) = 10 bits
MM address format

10 4 6
Virtual Memory:-
Virtual memory is the technique of using secondary storage such as disks to enter
the apparent size of accessible memory beyond its actual physical size. Virtual memory is
om
implemented by employing a memory-management unit (MMU) to translate every
logical address reference into a physical address reference as shown in fig 1. The MMU
c
is imposed between the CPU and the physical memory where it performs these
p.
translations under the control of the operating system. Each memory reference is sued by
the CPU is translated from the logical address space to the physical address space.
ou
Mapping tables guide the translation, again under the control of the operating system.
MMU
gr
CPU
ts
Mapping tastes
Cache Main Disk
- Memory
en
Virtual address
ud
Physical Address
Logical address
st
ity
Virtual memory usually demand paging, which means that a Page is moved from
disk into main memory only when the processor accesses a word on that page. Virtual
.c
memory pages always have a place on the disk once they are created, but are copied to
main memory only on a miss or page fault.
w
w
Advantages of Virtual memory:-

1. Simplified addressing:- Each program unit can be compiled into its own
w
memory space, beginning at address O and extending far beyond the limits of
physical memory. Programs and data structures do not require address relocation

at load time, nor must they be broken into fragments merely to accommodate
memory limitations.
2. Cost effective use of memory: - Less expensive disk storage can replace more
expensive RAM memory, since the entire program does not need to occupy
physical memory at one time.
om
3. Access control: - Since each memory reference must be translated, it can be
simultaneously checked for read, write and execute privileges. This allows
c
hardward level control of access to system resources and also prevents and also
p.
prevents buggy programs or intruders from causing damage to the resources of
other users or the system.
ou
a) Memory management by segmentation:- Segmentation allows memory to be
gr
divided into segments of varying sizes depending upon requirements. Fig 2. Shows a
main memory containing five segments identified by segment numbers. Each segment
ts
beings at a virtual address 0, regardless of where it is located in physical memory.
en
ud
st
ity
.c
w
w
w
Each virtual address arriving from the CPU is added to the contents of the
segment base register in the MMU to form the physical address. The virtual address may
also optionally be compared to a segment limit register to trap reference beyond a
specified limit.

b) Memory management by paging:- Fig 3 shows a simplified mechanism for virtual

address translation in a paged MMU. The process begins in a manner similar to the
segmentation process. The virtual address composed of a high order page number and a
low order word number is applied to MMU. The virtual page number is limit checked to
om
be certain that the page is within the page table, and if it is, it is added to the page table
base to yield the page table entry. The page table entry contains several control fields in
c
addition to the page field. The control fields may include access control bits, a presence
p.
bit, a dirty bit and one or more use bits, typically the access control field will include bits
specifying read, write and perhaps execute permission. The presence bit indicates
ou
whether the page is currently in main memory. The use bit is set upon a read or write to
the specified page, as an indication to the replaced algorithm in case a page must be
gr
replaced.
ts
en
ud
st
ity
.c
If the presence bit indicates a hit, then the page field of the page table entry will
w
contain the physical page number. If the presence bit is a miss, which is page fault, then
w
the page field of the page table entry which contains an address is secondary memory
where the page is stored. This miss condition also generates an interrupt. The interrupt
w
service routine will initiate the page fetch from secondary memory and with also
suspended the requesting process until the page has been bought into main memory. If

the CPU operation is a write hit, then the dirty bit is set. If the CPU operation is a write
miss, then the MMU with begin a write allocate process.
Problems:-
1. An address space is specified by 24 bits & the corresponding memory space is 16
bits.
om
a) How many words are there in address space?
b) How many words are there in memory space?
c
c) If a page has 2k words, how many pages & blocks are in the system?
p.
Solution:-
ou
a) Address space = 24 bits
224 = 24.220 = 16M words
gr
b) Memory space: 16 bits
216 = 64k words ts
c) page consists of 2k words
en
Number of pages in add space = 16M/2K = 8000
Number of blocks = 64k/2k = 32 blocks
ud
2. A virtual memory has a page size of 1k words. There are 8 pages & 4 blocks. The
associative memory page table has the following entries.
st
ity
Page block Make a list of virtual addresses (in decimal ) that will cause a page
0 3 by cpu fault of used
.c
1 1
w
w
4 2
w
6 0

The page numbers 2, 3, 5 & 7 are not available in associative memory page table
& hence cause a page fault.
5.7 PERIPHERAL DEVICES:
om
1. Magnetic Disk Drives: Hard disk Drive organization:
The modern hard disk drive is a system in itself. It contains not only the disks
c
that are used as the storage medium and the read write heads that access the raw data
p.
encoded on them, but also the signal conditioning circuitry and the interface electronics
that separate the system user from the details & getting bits on and off the magnetic
ou
surface. The drive has 4 platters with read/write heads on the top and bottom of each
platter. The drive rotates at a constant 3600rpm.
gr
Platters and Read/Write Heads: - The heart of the disk drive is the stack of rotating
ts
platters that contain the encoded data, and the read and write heads that access that data.
en
The drive contains five or more platters. There are read/write heads on the top and bottom
of each platter, so information can be recorded on both surfaces. All heads move together
ud
across the platters. The platters rotate at constant speed usually 3600 rpm.
Drive Electronics: - The disk drive electronics are located on a printed circuit board
st
attached to the disk drive. After a read request, the electronics must seek out and find the
ity
block requested, stream is off of the surface, error check and correct it, assembly into
bytes, store it in an on-board buffer and signal the processor that the task is complete.
.c
To assist in the task, the drive electronics include a disk controller, a special
w
purpose processor.
w
Data organization on the Disk:- The drive needs to know where the data to be accessed
w
is located on the disk. In order to provide that location information, data is organized on
the disk platters by tracks and sectors. Fig 13 shows simplified view of the organization
of tracks and sectors on a disk. The fig. Shows a disk with 1024 tracks, each of which has

64 sectors. The head can determine which track it is on by counting tracks from a known
location and sector identities are encoded in a header written on the disk at the front of
each sector.
The number of bytes per sector is fixed for a given disk drive, varying in size
om
from 512 bytes to 2KB. All tracks with the same number, but as different surfaces, form a
cylinder.
c
p.
The information is recorded on the disk surface 1 bit at a time by magnetizing a
small area on the track with the write head. That bit is detected by sending the direction
ou
of that magnetization as the magnetized area passes under the read head as shown in fig
14.
gr
ts
en
ud
st
ity
.c
Fig 5. shows how a typical sector might be organized. The header usually
w
contains both synchronization and location information. The synchronization information

w
allows the head positioning circuitry to keep the heads centered on the track and the
w
location information allows the disk controller to determine the sectors & identifies as the
header passes, so that the data can be captured if it is read or stored, if it is a write. The 12
bytes of ECC (Error Correcting Code) information are used to detect and correct errors in
the 512 byte data field.

Generally the disk drive manufacturer initializes or formats the drive by writing
its original track and sector information on the surfaces and checks to determine whether
data can be written and read from each sector. If any sectors are found to be bad, that is
incapable of being used even with ECC, then they are marked as being defective so their
om
use can be avoided by operating system.
c
The operating system interface:- The operating system specifies the track, sector and
p.
surface of the desired block. The disk controller translates that requests to a series of low
level disk operations. Fig 15 shows logical block address (LBA) used to communicate
ou
requests to the drive controller.
gr
Head #, Cylinder #, and Sector # refer to the starting address of the information
and sector count specifies the number of sectors requested.
ts
4 4 16 8 8
en
Drive # Head # Cylinder # Sector # Sector count
Logical block address
ud
Fig 15. Block address for disk address.

st
ity
The Disk Access Process:- Let us follow the process of reading a sector from the disk
1. The OS communicates the LBA to the disk drive and issues the read command.
.c
2. The drive seeks the correct track by moving the heads to the correct position and
enabling the one on the specified surface. The read feed reads sector numbers as
w
they travel by until the requested one is found.

w
3. Sector data and ECC stream into a buffer on the drive interface. ECC is done as
the fly.
w
4. The drive communicates data ready to the OS.

5. The OS reads data, either byte by byte or by issuing a DMA command.

Disk Drive Dynamic Properties: -
om
Dynamic properties are those that deal with the access time for the reading and
writing of data. The calculation of data access time is not simple. It depends not only as
c
the rotational speed of the disk, but also the location of the read/write head when it begins
p.
the access. There are several measures of data access times.
ou
1. Seek time: - Is the average time required to move the read/write head to the
desired track. Actual seek time which depend on where the head is when the
gr
request is received and how far it has to travel, but since there is no way to know
what these values will be when an access request is made, the average figure is
ts
used. Average seek time must be determined by measurement. It will depend on
en
the physical size of the drive components and how fast the heads can be
accelerated and decelerated. Seek times are generally in the range of 8-20 m sec
and have not changed much in recent years.
ud
2. Track to track access time: - Is the time required to move the head from one
track to adjoining one. This time is in the range of 1-2 m sec.
st
3. Rotational latency: - Is the average time required for the needed sector to pass
ity
under head once and head has been positioned once at the correct track. Since on
the average the desired sector will be half way around the track from where the
.c
head is when the head first arrives at the track, rotational latency is taken to be
the rotation time. Current rotation speeds are from 3600 to 7200 rpm, which yield
w
rotational latencies in the 4-8 ms range.

w
4. Average Access time:- Is equal to seek time plus rotational latency.

5. Burst rate: - Is the maximum rate at which the drive produces or accepts data
w
once the head reaches the desired sector, It is equal to the rate at which data bits
stream by the head, provided that the rest of the system can produce or accept data
at that rate

Burst rate (byte/sec) = rows/sec * sector/row * bytes/sector

6. Sustained data rate: - Is the rate at which data can be accessed over a sustained
period of time.
Problems 1: A hard drive has 8 surface, with 521 tracks per surface and a constant 64
om
sectors per track. Sector size is 1KB. The average seek time is 8 m sec, the track to track
access time is 1.5 m sec and the drive runs at 3600 rpm. Successive tracks in a cylinder
c
can be read without head movement.
p.
a) What is the drive capacity?
ou
b) What is the average access time for the drive?
c) Estimate the time required to transfer a 5MB file?
gr
d) What is the burst transfer rate?
Solution:- ts
a) Capacity = Surfaces x tracks x sectors x sector size
en
= 8 x 521 x 64 x 1k = 266.752MB
b) Rotational latency : Rotation time/2 = 60/3600x2 = 8.3 m sec
Average Access time = Seek time + Rotational latency
ud
= 8 m sec + 8.3 m sec = 16.3 m sec

c) Assume the file is stored in successive sectors and tracks starting at sector #0,
st
track#0 of cylinder #i. A 5MB file will need 1000 blocks and occupy from
ity
cylinder #1, track #0, sector #0 to cylinder #(i+9), track #6, sector #7, we also
assume the size of disk buffer is unlimited.
.c
The disk needs 8ms, which is the seek time, to find the cylinder #i, 8.3 ms to find
w
sector #0 and 8 x (60/3600) seconds to read all 8 tracks data of this cylinder. Then the
w
time needed for the read to move to next adjoining track will be only 1.5 m sec.
w
Which is the track to track access time. Assume a rotational latency before each new
track.

Access time = 8+9(8.38x16.6+1.5)+8.3+6x16.6+8/64x16.6

= 1406.6 m sec
d) Burst rate = rows/sec x sectors/row x bytes/sector
= 3600/60 x 64 x 1k = 3.84 MB/sec
om
Problem 2: A floppy disk drive has a maximum area bits density of 500 bits per mm2. its
innermost track is at a radius of 2cm, and its outermost track is at radius of 5cm. It rotates
c
at 300rpm.
p.
a) What is the total capacity of the drive?
ou
b) What is its burst data rate in bytes per second?
Solution:-
gr
a) At the maximum area density the bit spacing equals the track spacing t, so
ts
500 = 1/t2
en
or t = 0.045 mm
The number of tracks = (50-20)/0.045 = 666
The number of bits per track = 40p/0.045 2791
ud
Thus the unformatted or raw capacity is 666 x 2791 = 1,841,400 bits

b) Given 5 raw per second the (raw)burst rate is
st
Rev/second x bytes/rev = 5 x 2791/8 = 1744 bytes/sec

ity
Optical Disks
.c
Compact Disk (CD) Technology:- The optical technology that is used for CD system is
based on laser light source. A laser beam is directed onto the surface of the spinning disk.
w
Physical indentations in the surface are arranged along the tracks of the disk. They reflect
w
the focused beam towards a photo detector, which detects the stored binary patterns.
The laser emits a coherent light beam that is sharply focused on the surface of the
w
disk. Coherent light consists of Synchronized waves that have the same wavelength. If a
coherent light beam is combined with another beam of the same kind, and the two beams
are in phase, then the result will be brighter beam. But, if a photo detector is used to

detect the beams, it will detect a bright spot in the first case and a dark spot in the second
case.
A cross section of a small portion of a CD shown in fig. 16
The bottom payer is Polycarbonate plastic, which functions as a clear glass base. The
surface of this plastic is Programmed to store data by indenting it with pits. The
om
unindented parts are called lands. A thin layer of reflecting aluminium material is placed
on top of a programmed disk. The aluminium is then covered by a protective acrylic.
c
Finally the topmost layer is deposited and stamped with a label.
p.
ou
gr
ts
en
The laser source and the Photo detector are positioned below the polycarbonate
ud
plastic. The emitted bean travels through this plastic, reflects off the aluminium layer and
travels back toward photo detector.
st
Some important optical disks are listed below

ity
1. CD-ROM
2. CD-RWs (CD-re writables)
.c
3. DVD technology (Digital Versatile disk)

w
Problem 1:- A certain two-way associative cache has an access time of 40 nano seconds,
w
compared to a miss time of 90 nano seconds. Without the cache, main memory access
w
time was 90 nano seconds. Without the cache, main memory access time was 70 nano
seconds. Running a set of benchmarks with and without the cache indicated a speed up of
1.4.
What is the approximate hit rate?

Solution:- Hit:tn = 40 nano seconds

Miss:tm = 90
No Cache: tn = 70
Access time with Cache ta = h.tn+(1-h)tm = tn-h(tn-tn)
om
Speed up S = tn/ta = 70/90-50h = 1.4
Which gives hit ratio h = 80%
c
p.
Problem 2: A 16 MB main memory has a 32 KB direct mapped Cache with 8 bytes per
line.
ou
a. How many lines are there in the Cache.
gr
b. Show how the main memory address is Partitioned?
ts
Solution:- a. Number of Cache lines = 215/23 = 212 lines
en
b. 23 15 14 32 0
Tag (9 bits) Group (12 bits) Bytes (3)
16MB = 24.220 = 224
ud
st
ity
.c
w
w
w

UNIT - 6
Arithmetic: Addition and Subtraction of Signed Numbers, Design of Fast
Adders, Multiplication of Positive Numbers, Signed Operand Multiplication,
om
Fast Multiplication, Integer Division, Floating-point Numbers and
Operations
c
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
w

UNIT - 6
ARITHMETIC
6.1 ADDITION AND SUBTRACTION OF SIGNED NUMBERS:
om
In figure-1, the function table of a full-adder is shown; sum and carryout are the outputs
for adding equally weighted bits xi and yi, in two numbers X and Y. The logic
expressions for these functions are also shown, along with an example of addition of the
c
4-bit unsigned numbers 7 and 6. Note that each stage of the addition process must
p.
accommodate a carry-in bit. We use ci, to represent the carry-in to the ith stage, which is
ou
the same as the carryout from the (i - 1) th stage.
The logic expression for si in Figure-1 can be implemented with a 3-input XOR gate. The
carryout function, ci is implemented with a two-level AND-OR logic circuit. A
gr
+1
convenient symbol for the complete circuit for a single stage of addition, called a full
ts
adder (FA), is as shown in the figure-1a.
A cascaded connection of such n full adder blocks, as shown in Figure 1b, forms a
en
parallel adder & can be used to add two n-bit numbers. Since the carries must propagate,
or ripple, through this cascade, the configuration is called an n-bit ripple-carry adder.
ud
The carry-in, Co, into the least-significant-bit (LSB) position [Ist stage] provides a
convenient means of adding 1 to a number. Take for instance; forming the 2's-
st
complement of a number involves adding 1 to the 1s-complement of the number. The

carry signals are also useful for interconnecting k adders to form an adder capable of
ity
handling input numbers that are kn bits long, as shown in Figure-1c.

.c
w
w
w

c om
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
FIG-1: Addition of binary vectors.

w

c om
p.
ou
FIG-2: 4 - Bit parallel Adder.
gr
6.2 DESIGN OF FAST ADDERS: ts
In an n-bit parallel adder (ripple-carry adder), there is too much delay in
en
developing the outputs, so through sn-1 and cn. On many occasions this delay is not
acceptable; in comparison with the speed of other processor components and speed of the
data transfer between registers and cache memories. The delay through a network
ud
depends on the integrated circuit technology used in fabricating the network and on the
number of gates in the paths from inputs to outputs (propagation delay). The delay
st
through any combinational logic network constructed from gates in a particular

ity
technology is determined by adding up the number of logic-gate delays along the longest
signal propagation path through the network. In the case of the n-bit ripple-carry adder,
.c
the longest path is from inputs x0, y0, and c0 at the least-significant-bit (LSB) position to
outputs cn and sn-1 at the most-significant-bit (MSB) position.
w
Using the logic implementation indicated in Figure-1, cn-1 is available in 2(n1)

w
gate delays, and sn-1 is one XOR gate delay later. The final carry-out, cn is available after
2n gate delays. Therefore, if a ripple-carry adder is used to implement the
w
addition/subtraction unit shown in Figure-3, all sum bits are available in 2n gate delays,
including the delay through the XOR gates on the Y input. Using the implementation

cn cn-1 for overflow, this indicator is available after 2n+2 gate delays. In summary, in a

parallel adder an nth stage adder can not complete the addition process before all its
previous stages have completed the addition even with input bits ready. This is because,
the carry bit from previous stage has to be made available for addition of the present
stage.
om
In practice, a number of design techniques have been used to implement high-
speed adders. In order to reduce this delay in adders, an augmented logic gate network
c
structure may be used. One such method is to use circuit designs for fast propagation of
p.
carry signals (carry prediction).
ou
Carry-Look ahead Addition:
gr
As it is clear from the previous discussion that a parallel adder is considerably
slow & a fast adder circuit must speed up the generation of the carry signals, it is
ts
necessary to make the carry input to each stage readily available along with the input bits.
This can be achieved either by propagating the previous carry or by generating a carry
en
depending on the input bits & previous carry. The logic expressions for si (sum) and c i+1
(carry-out) of stage ith are
ud
st
ity
.c
w
w
w

The above expressions Gi and Pi are called carry generate and propagate functions
for stage i. If the generate function for stage i is equal to 1, then ci+1 = 1, independent of
the input carry, ci. This occurs when both xi and yi are 1. The propagate function means
that an input carry will produce an output carry when either xi or yi or both equal to 1.
Now, using Gi & Pi functions we can decide carry for ith stage even before its previous
om
stages have completed their addition operations. All Gi and Pi functions can be formed
independently and in parallel in only one gate delay after the Xi and Yi inputs are applied
c
to an n-bit adder. Each bit stage contains an AND gate to form Gi, an OR gate to form Pi
p.
and a three-input XOR gate to form si. However, a much simpler circuit can be derived
by considering the propagate function as Pi = xi yi, which differs from Pi = xi + yi only
ou
when xi = yi =1 where Gi = 1 (so it does not matter whether Pi is 0 or 1). Then, the basic
diagram in Figure-5 can be used in each bit stage to predict carry ahead of any stage
gr
completing its addition.
Consider the ci+1expression, ts
en
This is because, Ci = (Gi-1 + Pi-1Ci-1).
Further, Ci-1 = (Gi-2 + Pi-2Ci-2) and so on. Expanding in this fashion, the final carry
ud
expression can be written as below;

st
C i+1 = Gi + PiG i-1 + PiP i-1 G i-2 + + Pi P i-1 P 1G0 + Pi P i-1 P0G0
ity
.c
Thus, all carries can be obtained in three gate delays after the input signals Xi, Yi
and Cin are applied at the inputs. This is because only one gate delay is needed to
w
develop all Pi and Gi signals, followed by two gate delays in the AND-OR circuit (SOP
w
expression) for ci + 1. After a further XOR gate delay, all sum bits are available.
Therefore, independent of n, the number of stages, the n-bit addition process requires
w
only four gate delays.

c om
p.
ou
gr
ts
en
FIG-5: 4 bit carry look ahead adder.

ud
Now, consider the design of a 4-bit parallel adder. The carries can be implemented as
st
;i = 0
;i = 1
ity
;i = 2
;i = 3
.c
w
The complete 4-bit adder is shown in Figure 5b where the B cell indicates Gi, Pi & Si
w
generator. The carries are implemented in the block labeled carry look-ahead logic. An
adder implemented in this form is called a carry look ahead adder. Delay through the
w
adder is 3 gate delays for all carry bits and 4 gate delays for all sum bits. In comparison,
note that a 4-bit ripple-carry adder requires 7 gate delays for S3(2n-1) and 8 gate
delays(2n) for c4.

If we try to extend the carry look- ahead adder of Figure 5b for longer operands,
we run into a problem of gate fan-in constraints. From the final expression for Ci+1 & the
carry expressions for a 4 bit adder, we see that the last AND gate and the OR gate require
a fan-in of i + 2 in generating cn-1. For c4 (i = 3)in the 4-bit adder, a fan-in of 5 is required.
om
This puts the limit on the practical implementation. So the adder design shown in Figure
4b cannot be directly extended to longer operand sizes. However, if we cascade a number
c
of 4-bit adders, it is possible to build longer adders without the practical problems of fan-
p.
in. An example of a 16 bit carry look ahead adder is as shown in figure 6. Eight 4-bit
carry look-ahead adders can be connected as in Figure-2 to form a 32-bit adder.
ou
gr
ts
en
ud
st
ity
FIG-6: 16 bit carry-look ahead adder.

.c
6.3 MULTIPLICATION OF POSITIVE NUMBERS:

w
Consider the multiplication of two integers as in Figure-6a in binary number

system. This algorithm applies to unsigned numbers and to positive signed numbers. The
w
product of two n-digit numbers can be accommodated in 2n digits, so the product of the
w
two 4-bit numbers in this example fits into 8 bits. In the binary system, multiplication by
the multiplier bit 1 means the multiplicand is entered in the appropriate position to be
added to the partial product. If the multiplier bit is 0, then 0s are entered, as indicated
in the third row of the shown example.

Binary multiplication of positive operands can be implemented in a combinational

(speed up) two-dimensional logic array, as shown in Figure 7. Here, M- indicates
multiplicand, Q- indicates multiplier & P- indicates partial product. The basic
om
component in each cell is a full adder FA. The AND gate in each cell determines whether
a multiplicand bit mj, is added to the incoming partial-product bit, based on the value of
c
the multiplier bit, qi. For i in the range of 0 to 3, if qi = 1, add the multiplicand
p.
(appropriately shifted) to the incoming partial product, PPi, to generate the outgoing
partial product, PP(i+ 1) & if qi = 0, PPi is passed vertically downward unchanged. The
ou
initial partial product PPO is all 0s. PP4 is the desired product. The multiplicand is
shifted left one position per row by the diagonal signal path. Since the multiplicand is
gr
shifted and added to the partial product depending on the multiplier bit, the method is
referred as SHIFT & ADD method. The multiplier array & the components of each bit
ts
cell are indicated in the diagram, while the flow diagram shown explains the
en
multiplication procedure.
ud
st
ity
.c
w
w
w
FIG-7a
P7, P6, P5,,P0 product.

c om
p.
FIG-7b
ou
The following SHIFT & ADD method flow chart depicts the multiplication logic for
unsigned numbers.
gr
ts
en
ud
st
ity
.c
w
w
w

c om
p.
ou
gr
ts
en
ud
st
Despite the use of a combinational network, there is a considerable amount of

ity
delay associated with the arrangement shown. Although the preceding combinational
multiplier is easy to understand, it uses many gates for multiplying numbers of practical
.c
size, such as 32- or 64-bit numbers. The worst case signal propagation delay path is from
the upper right corner of the array to the high-order product bit output at the bottom left
w
corner of the array. The path includes the two cells at the right end of each row, followed
w
by all the cells in the bottom row. Assuming that there are two gate delays from the
w
inputs to the outputs of a full adder block, the path has a total of 6(n - 1) - 1 gate delays,
including the initial AND gate delay in all cells, for the n x n array. In the delay
expression, (n-1) because, only the AND gates are actually needed in the first row of the
array because the incoming (initial) partial product PPO is zero.

Multiplication can also be performed using a mixture of combinational array

techniques (similar to those shown in Figure 7) and sequential techniques requiring less
combinational logic. Multiplication is usually provided as an instruction in the machine
instruction set of a processor. High-performance processor (DS processors) chips use an
appreciable area of the chip to perform arithmetic functions on both integer and floating-
om
point operands. Sacrificing an area on-chip for these arithmetic circuits increases the
speed of processing. Generally, processors built for real time applications have an on-
c
chip multiplier.
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
w

c om
p.
ou
gr
ts
en
ud
st
ity
.c
w
FIG-8a
w
w
Another simplest way to perform multiplication is to use the adder circuitry in the
ALU for a number of sequential steps. The block diagram in Figure 8a shows the
hardware arrangement for sequential multiplication. This circuit performs multiplication
by using single n-bit adder n times to implement the spatial addition performed by the n

rows of ripple-carry adders. Registers A and Q combined to hold PPi while multiplier bit
qi generates the signal Add/No-add. This signal controls the addition of the multiplicand
M to PPi to generate PP(i + 1). The product is computed in n cycles. The partial product
grows in length by one bit per cycle from the initial vector, PPO, of n 0s in register A.
The carry-out from the adder is stored in flip-flop C. To begin with, the multiplier is
om
loaded into register Q, the multiplicand into register M and registers C and A are cleared
to 0. At the end of each cycle C, A, and Q are shifted right one bit position to allow for
c
growth of the partial product as the multiplier is shifted out of register Q. Because of this
p.
shifting, multiplier bit qi, appears at the LSB position of Q to generate the Add/No-add
signal at the correct time, starting with qo during the first cycle, q1 during the second
ou
cycle, and so on. After they are used, the multiplier bits are discarded by the right-shift
operation. Note that the carry-out from the adder is the leftmost bit of PP(i + 1), and it
gr
must be held in the C flip-flop to be shifted right with the contents of A and Q. After n
cycles, the high-order half-of- the product is held in register A and the low-order half is
ts
in register Q. The multiplication example used above is shown in Figure 8b as it would
en
be performed by this hardware arrangement.
M
ud
1 10 1
Initial
st
0 0 00 0 1 0 1 1 Configuration
C A Q
ity
0 1 1 0 1 1 0 1 1 Add I cycle
0 0 1 1 0 1 1 0 1 Shift
.c
1 0 0 1 1 1 1 0 1 Add II cycle
0 1 0 0 1 1 1 1 0 Shift
w
w
0 1 0 0 1 1 1 1 0 No Add III cycle

0 0 1 0 0 1 1 1 1 Shift
w
1 0 0 0 1 1 1 1 1 Add IV cycle
0 1 0 0 0 1 1 1 1 Shift
Product FIG 8b

FIG-7b
Using this sequential hardware structure, it is clear that a multiply instruction takes
much more time to execute than an Add instruction. This is because of the sequential
circuits associated in a multiplier arrangement. Several techniques have been used to
speed up multiplication; bit pair recoding, carry save addition, repeated addition, etc.
om
6.4 SIGNED-OPERAND MULTIPLIATION:
c
p.
Multiplication of 2's-complement signed operands, generating a double-length
ou
product is still achieved by accumulating partial products by adding versions of the
multiplicand as decided by the multiplier bits. First, consider the case of a positive
gr
multiplier and a negative multiplicand. When we add a negative multiplicand to a
partial product, we must extend the sign-bit value of the multiplicand to the left as far as
ts
the product will extend. In Figure 9, for example, the 5-bit signed operand, - 13, is the
multiplicand, and +11, is the 5 bit multiplier & the expected product -143 is 10-bit wide.
en
The sign extension of the multiplicand is shown in red color. Thus, the hardware
discussed earlier can be used for negative multiplicands if it provides for sign extension
ud
of the partial products.

st
0 0 1 1 (-13) X 0 1 0 1 1 (+11)
ity
111111 0 0 1 1
111110 0 1 1
000000 0 0
.c
111001 1
000000
w
110111 0 0 0 1 (-143)
w
FIG-9
w
For a negative multiplier, a straightforward solution is to form the 2's-

complement of both the multiplier and the multiplicand and proceed as in the case of a
positive multiplier. This is possible because complementation of both operands does not

change the value or the sign of the product. In order to take care of both negative and
positive multipliers, BOOTH algorithm can be used.
Booth Algorithm
om
The Booth algorithm generates a 2n-bit product and both positive and negative
2's-complement n-bit operands are uniformly treated. To understand this algorithm,
c
consider a multiplication operation in which the multiplier is positive and has a single
p.
block of 1s, for example, 0011110(+30). To derive the product, as in the normal standard
procedure, we could add four appropriately shifted versions of the multiplicand,.
ou
However, using the Booth algorithm, we can reduce the number of required operations by
regarding this multiplier as the difference between numbers 32 & 2 as shown below;
gr
0 1 0 0 0 0 0 (32)
0 0 0 0 0 1 0 (-2)
ts
0 0 1 1 1 1 0 (30)
en
This suggests that the product can be generated by adding 25 times the
ud
multiplicand to the 2's-complement of 21 times the multiplicand. For convenience, we can

describe the sequence of required operations by recoding the preceding multiplier as 0
st
+1000 - 10. In general, in the Booth scheme, -1 times the shifted multiplicand is selected
ity
when moving from 0 to 1, and +1 times the shifted multiplicand is selected when moving
from 1 to 0, as the multiplier is scanned from right to left.
.c
0 1 0 1 1 0 1
w
0 0+1 +1 +1 +1 0
0 0 0 0 0 0 0
w
0 1 0 1 1 0 1
0 1 0 1 1 0 1
w
0 1 0 1 1 0 1
01 0 1 1 0 1
0 00 0 0 0 0
00 00 0 0 0
00 010 1 0 1 0 0 01 1 0

FIG-10a: Normal Multiplication
0 1 0 1 1 0 1
om
0 0+1 +1 +1 +1 0
0 0 0 0 0 0 00 0 0 0 0 0 0
1 1 1 1 1 1 1 01 0 0 1 1
c
0 0 0 0 0 0 0 00 0 0 0
0 0 0 0 0 0 0 00 0 0
p.
0 0 0 0 0 0 0 00 0
0 0 0 1 0 1 1 01
ou
0 0 0 0 0 0 0 0
0 0 0 1 0 1 0 10 0 0 1 1 0
gr
FIG-10b: Booth Multiplication
Figure 10 illustrates the normal and the Booth algorithms for the said example.
ts
The Booth algorithm clearly extends to any number of blocks of 1s in a multiplier,
en
including the situation in which a single 1 is considered a block. See Figure 11a for
another example of recoding a multiplier. The case when the least significant bit of the
multiplier is 1 is handled by assuming that an implied 0 lies to its right. The Booth
ud
algorithm can also be used directly for negative multipliers, as shown in Figure 11a.
To verify the correctness of the Booth algorithm for negative multipliers, we use the
st
following property of negative-number representations in the 2's-complement

ity
.c
w
w
w
FIG-11a

c om
p.
ou
gr
FIG-11b
6.5 FAST MULIPLICATION:

ts
en
There are two techniques for speeding up the multiplication operation. The first
technique guarantees that the maximum number of summands (versions of the multiplicand)
ud
that must be added is n/2 for n-bit operands. The second technique reduces the time needed
to add the summands (carry-save addition of summands method).
st
Bit-Pair Recoding of Multipliers:

ity
This bit-pair recoding technique halves the maximum number of summands. It is

derived from the Booth algorithm. Group the Booth-recoded multiplier bits in pairs, and
.c
observe the following: The pair (+1 -1) is equivalent to the pair (0 +1). That is, instead of
adding 1 times the multiplicand M at shift position i to + 1 x M at position i + 1, the
w
same result is obtained by adding +1 x M at position I Other examples are: (+1 0) is

w
equivalent to (0 +2),(-l +1) is equivalent to (0 1). and so on. Thus, if the Booth-recoded
w
multiplier is examined two bits at a time, starting from the right, it can be rewritten in a form
that requires at most one version of the multiplicand to be added to the partial product for
each pair of multiplier bits. Figure 14a shows an example of bit-pair recoding of the
multiplier in Figure 11, and Figure 14b shows a table of the multiplicand

c om
p.
ou
gr
ts
en
ud
st
ity
FIG - 14
.c
selection decisions for all possibilities. The multiplication operation in figure 11a is
shown in Figure 15. It is clear from the example that the bit pair recoding method
w
requires only n/2 summands as against n summands in Booths algorithm.

w
w

c om
p.
ou
gr
ts
en
ud
st
ity
FIG 15: Multiplication requiring n/2 summands.

.c
w
6.6 INTEGER DIVISION:

w
Positive-number multiplication operation is done manually in the way it is done in a

w
logic circuit. A similar kind of approach can be used here in discussing integer division.
First, consider positive-number division. Figure 16 shows examples of decimal division
and its binary form of division. First, let us try to divide 2 by13, and it does not work.
Next, let us try to divide 27 by 13. Going through the trials, we enter 2 as the quotient

and perform the required subtraction. The next digit of the dividend, 4, is brought down,
and we finish by deciding that 13 goes into 14 once and the remainder is 1. Binary
division is similar to this, with the quotient bits only 0 and 1.
A circuit that implements division by this longhand method operates as follows: It
positions the divisor appropriately with respect to the dividend and performs a subtrac-
om
tion. If the remainder is zero or positive, a quotient bit of 1 is determined, the remainder
is extended by another bit of the dividend, the divisor is repositioned, and subtraction is
c
performed. On the other hand, if the remainder is negative, a quotient bit of 0 is
p.
determined, the dividend is restored by adding back the divisor, and the divisor H
repositioned for another subtraction
ou
gr
ts
en
ud
st
ity
.c
FIG - 16
w
w
w

c om
p.
ou
gr
ts
en
ud
FIG 17: Binary Division

st
ity
.c
w
w
w

c om
p.
ou
gr
ts
FIG 18: Restoring Division
en
ud
Restoring Division:
Figure 17 shows a logic circuit arrangement that implements restoring division. Note its
st
similarity to the structure for multiplication that was shown in Figure 8. An n-bit positive
ity
divisor is loaded into register M and an n-bit positive dividend is loaded into register Q at
the start of the operation. Register A is set to 0. After the division is complete, the n-bit
.c
quotient is in register Q and the remainder is in register A. The required subtractions are
facilitated by using 2's-complement arithmetic. The extra bit position at the left end of
w
both A and M accommodates the sign bit during subtractions. The following algorithm
w
performs restoring division.

Do the following n times:
w
1. Shift A and Q left one binary position.

2. Subtract M from A, and place the answer back in A.

3. If the sign of A is 1, set q0 to 0 and add M back to A (that is, restore A);
otherwise, set q0to 1.
Figure 18 shows a 4-bit example as it would be processed by the circuit in Figure 17.
om
No restoring Division:
The restoring-division algorithm can be improved by avoiding the need for restoring A
c
after an unsuccessful subtraction. Subtraction is said to be unsuccessful if the result
p.
ou
is negative. Consider the sequence of operations that takes place after the subtraction
operation in the preceding algorithm. If A is positive, we shift left and subtract M, that is,
gr
we perform 2A - M. If A is negative, we restore it by performing A + M, and then we
shift it left and subtract M. This is equivalent to performing 2A + M. The q0 bit is
ts
appropriately set to 0 or 1 after the correct operation has been performed. We can
summarize this in the following algorithm for no restoring division.
en
Step 1: Do the following n times:
1.If the sign of A is 0, shift A and Q left one bit position and subtract M from
ud
A; otherwise, shift A and Q left and add M to A.

2. Now, if the sign of A is 0, set q0 to 1; otherwise, set q0 to 0.
st
Step 2: If the sign of A is 1, add M to A.

ity
Step 2 is needed to leave the proper positive remainder in A at the end of the n
cycles of Step 1. The logic circuitry in Figure 17 can also be used to perform this
.c
algorithm. Note that the Restore operations are no longer needed, and that exactly one
w
Add or Subtract operation is performed per cycle. Figure 19 shows how the division
example in Figure 18 is executed by the no restoring-division algorithm.
w
There are no simple algorithms for directly performing division on signed operands that
w
are comparable to the algorithms for signed multiplication. In division, the operands can
be preprocessed to transform them into positive values. After using one of the algorithms

just discussed, the results are transformed to the correct signed values, as necessary.
c om
p.
ou
gr
ts
en
FIG 19: Non-restoring Division
ud
6.7 FLOATING-POINT NUMBERS AND OPERATIONS:

st
Floating point arithmetic is an automatic way to keep track of the radix point.
ity
The discussion so far was exclusively with fixed-point numbers which are considered as
.c
integers, that is, as having an implied binary point at the right end of the number. It is
also possible to assume that the binary point is just to the right of the sign bit, thus
w
representing a fraction or any where else resulting in real numbers. In the 2's-complement
w
system, the signed value F, represented by the n-bit binary fraction

B = b0.b - 1b -2 ..b-(n-1) is given by
w
F(B) = -bo x 2 + b-1 x 2-1 +b-2x2-2 + ... + b-(n-X) x 2-{n~l) where the range of F is
-1 F 1 -2-(n-1). Consider the range of values representable in a 32-bit, signed, fixed-
point format. Interpreted as integers, the value range is approximately 0 to 2.15 x 109. If

we consider them to be fractions, the range is approximately 4.55 x 10-10 to 1. Neither

of these ranges is sufficient for scientific calculations, which might involve parameters
like Avogadro's number (6.0247 x 1023 mole-1) or Planck's constant (6.6254 x 10-27erg s).
Hence, we need to easily accommodate both very large integers and very small fractions.
To do this, a computer must be able to represent numbers and operate on them in such a
om
way that the position of the binary point is variable and is automatically adjusted as
computation proceeds. In such a case, the binary point is said to float, and the numbers
c
are called floating-point numbers. This distinguishes them from fixed-point numbers,
p.
whose binary point is always in the same position.
Because the position of the binary point in a floating-point number is variable, it
ou
must be given explicitly in the floating-point representation. For example, in the familiar
decimal scientific notation, numbers may be written as 6.0247 x 1023, 6.6254 -10-27, -
gr
1.0341 x 102, -7.3000 x 10-14, and so on. These numbers are said to be given to five
significant digits. The scale factors (1023, 10-27, and so on) indicate the position of the
ts
decimal point with respect to the significant digits. By convention, when the decimal
en
point is placed to the right of the first (nonzero) significant digit, the number is said to be
normalized. Note that the base, 10, in the scale factor is fixed and does not need to appear
explicitly in the machine representation of a floating-point number. The sign, the
ud
significant digits, and the exponent in the scale factor constitute the representation. We
are thus motivated to define a floating-point number representation as one in which a
st
number is represented by its sign, a string of significant digits, commonly called the
ity
mantissa, and an exponent to an implied base for the scale factor.

.c
w
w
w

UNIT - 7
Basic Processing Unit: Some Fundamental Concepts, Execution of a
Complete Instruction, Multiple Bus Organization, Hard-wired Control,
om
Microprogrammed Control
c
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
w

UNIT - 7
BASIC PROCESSING UNIT
BASIC PROCESSING UNIT:
om
The heart of any computer is the central processing unit (CPU). The CPU
executes all the machine instructions and coordinates the activities of all other units
c
during the execution of an instruction. This unit is also called as the Instruction Set
p.
Processor (ISP). By looking at its internal structure, we can understand how it performs
ou
the tasks of fetching, decoding, and executing instructions of a program. The processor is
generally called as the central processing unit (CPU) or micro processing unit (MPU).An
high-performance processor can be built by making various functional units operate in
gr
parallel. High-performance processors have a pipelined organization where the execution
ts
of one instruction is started before the execution of the preceding instruction is
completed. In another approach, known as superscalar operation, several instructions are
en
fetched and executed at the same time. Pipelining and superscalar architectures provide a
very high performance for any processor.
ud
A typical computing task consists of a series of steps specified by a sequence of

machine instructions that constitute a program. A program is a set of instructions
st
performing a meaningful task. An instruction is command to the processor & is executed

by carrying out a sequence of sub-operations called as micro-operations. Figure 1
ity
indicates various blocks of a typical processing unit. It consists of PC, IR, ID, MAR,
MDR, a set of register arrays for temporary storage, Timing and Control unit as main
.c
units.
w
w
7.1 FUNDAMENTAL CONCEPTS:

w
Execution of a program by the processor starts with the fetching of instructions

one at a time, decoding the instruction and performing the operations specified. From
memory, instructions are fetched from successive locations until a branch or a jump

instruction is encountered. The processor keeps track of the address of the memory
location containing the next instruction to be fetched using the program counter (PC) or
Instruction Pointer (IP). After fetching an instruction, the contents of the PC are updated
to point to the next instruction in the sequence. But, when a branch instruction is to be
executed, the PC will be loaded with a different (jump/branch address).
c om
p.
ou
gr
ts
en
Fig-1
ud
Instruction register, IR is another key register in the processor, which is used to

hold the op-codes before decoding. IR contents are then transferred to an instruction
st
decoder (ID) for decoding. The decoder then informs the control unit about the task to be
executed. The control unit along with the timing unit generates all necessary control
ity
signals needed for the instruction execution. Suppose that each instruction comprises 2
bytes, and that it is stored in one memory word. To execute an instruction, the processor
.c
has to perform the following three steps:

w
1. Fetch the contents of the memory location pointed to by the PC. The contents
w
of this location are interpreted as an instruction code to be executed. Hence, they are
w
loaded into the IR/ID. Symbolically, this operation can be written as

IR [(PC)]
2. Assuming that the memory is byte addressable, increment the contents of the
PC by 2, that is,

PC [PC] + 2
3. Decode the instruction to understand the operation & generate the control
signals necessary to carry out the operation.
4. Carry out the actions specified by the instruction in the IR.
om
In cases where an instruction occupies more than one word, steps 1 and 2 must be
c
repeated as many times as necessary to fetch the complete instruction. These two steps
p.
together are usually referred to as the fetch phase; step 3 constitutes the decoding phase;
and step 4 constitutes the execution phase.
ou
To study these operations in detail, let us examine the internal organization of the
processor. The main building blocks of a processor are interconnected in a variety of
gr
ways. A very simple organization is shown in Figure 2. A more complex structure that
provides high performance will be presented at the end.
ts
en
ud
st
ity
.c
w
w
w
Fig 2

Figure shows an organization in which the arithmetic and logic unit (ALU) and all
the registers are interconnected through a single common bus, which is internal to the
processor. The data and address lines of the external memory bus are shown in Figure 7.1
connected to the internal processor bus via the memory data register, MDR, and the
om
memory address register, MAR, respectively. Register MDR has two inputs and two
outputs. Data may be loaded into MDR either from the memory bus or from the internal
c
processor bus. The data stored in MDR may be placed on either bus. The input of MAR
p.
is connected to the internal bus, and its output is connected to the external bus. The
control lines of the memory bus are connected to the instruction decoder and control logic
ou
block. This unit is responsible for issuing the signals that control the operation of all the
units inside the processor and for interacting with the memory bus.
gr
The number and use of the processor registers RO through R(n - 1) vary considerably
from one processor to another. Registers may be provided for general-purpose use by the
ts
programmer. Some may be dedicated as special-purpose registers, such as index registers
en
or stack pointers. Three registers, Y, Z, and TEMP in Figure 2, have not been mentioned
before. These registers are transparent to the programmer, that is, the programmer need
not be concerned with them because they are never referenced explicitly by any
ud
instruction. They are used by the processor for temporary storage during execution of
some instructions. These registers are never used for storing data generated by one
st
instruction for later use by another instruction.

ity
The multiplexer MUX selects either the output of register Y or a constant value 4 to be
provided as input A of the ALU. The constant 4 is used to increment the contents of the
.c
program counter. We will refer to the two possible values of the MUX control input
Select as Select4 and Select Y for selecting the constant 4 or register Y, respectively.
w
As instruction execution progresses, data are transferred from one register to another,
w
often passing through the ALU to perform some arithmetic or logic operation. The
instruction decoder and control logic unit is responsible for implementing the actions
w
specified by the instruction loaded in the IR register. The decoder generates the control
signals needed to select the registers involved and direct the transfer of data. The

registers, the ALU, and the interconnecting bus are collectively referred to as the data
path.
With few exceptions, an instruction can be executed by performing one or more of the
following operations in some specified sequence:
1. Transfer a word of data from one processor register to another or to the ALU
om
2. Perform an arithmetic or a logic operation and store the result in a processor
register
c
3. Fetch the contents of a given memory location and load them into a processor
p.
register
4. Store a word of data from a processor register into a given memory location
ou
We now consider in detail how each of these operations is implemented, using the simple
processor model in Figure 2.
gr
Instruction execution involves a sequence of steps in which data are transferred from one
ts
register to another. For each register, two control signals are used to place the contents of
en
that register on the bus or to load the data on the bus into the register. This is represented
symbolically in Figure 3. The input and output of register Ri are connected to the bus via
switches controlled by the signals Riin and Riout respectively. When Riin is set to 1, the
ud
data on the bus are loaded into Ri. Similarly, when Riout, is set to 1, the contents of
register Riout are placed on the bus. While Riout is equal to 0, the bus can be used for
st
transferring data from other registers.

ity
Suppose that we wish to transfer the contents of register RI to register R4. This can be
accomplished as follows:
.c
1. Enable the output of register R1out by setting Rlout, tc 1. This places the contents
of R1 on the processor bus.
w
2. Enable the input of register R4 by setting R4in to 1. This loads data from the
w
processor bus into register R4.

All operations and data transfers within the processor take place within time periods
w
defined by the processor clock. The control signals that govern a particular transfer are
asserted at the start of the clock cycle. In our example, Rlout and R4in are set to 1. The
registers consist of edge-triggered flip-flops. Hence, at the next active edge of the clock,

the flip-flops that constitute R4 will load the data present at their inputs. At the same
time, the control signals Rlout and R4in will return to 0. We will use this simple model of
the timing of data transfers for the rest of this chapter. However, we should point out that
other schemes are possible. For example, data transfers may use both the rising and
falling edges of the clock. Also, when edge-triggered flip-flops are not used, two or more
om
clock signals may be needed to guarantee proper transfer of data. This is known as
multiphase clocking.
c
An implementation for one bit of register Ri is shown in Figure 7.3 as an example. A
p.
two-input multiplexer is used to select the data applied to the input of an edge-triggered
D flip-flop. When the control input Riin is equal to 1, the multiplexer selects the data on
ou
the bus. This data will be loaded into the flip-flop at the rising edge of the clock. When
Riin is equal to 0, the multiplexer feeds back the value currently stored in the flip-flop.
gr
The Q output of the flip-flop is connected to the bus via a tri-state gate. When Riout, is
equal to 0, the gate's output is in the high-impedance (electrically disconnected) state.
ts
This corresponds to the open-circuit state of a switch. When Riout, = 1, the gate drives the
en
bus to 0 or 1, depending on the value of Q.
ud
7.2 EXECUTION OF A COMPLETE INSTRUCTION:

st
Let us now put together the sequence of elementary operations required to execute one
ity
instruction. Consider the instruction

Add (R3), R1
.c
which adds the contents of a memory location pointed to by R3 to register R1. Executing
this instruction requires the following actions:
w
1. Fetch the instruction.

w
2. Fetch the first operand (the contents of the memory location pointed to by R3).
3. Perform the addition.
w
4 .Load the result into Rl.

c om
p.
Fig 7
ou
The listing shown in figure 7 above indicates the sequence of control steps
required to perform these operations for the single-bus architecture of Figure 2.
gr
Instruction execution proceeds as follows. In step 1, the instruction fetch operation is
initiated by loading the contents of the PC into the MAR and sending a Read request to
ts
the memory. The Select signal is set to Select4, which causes the multiplexer MUX to
en
select the constant 4. This value is added to the operand at input B, which is the contents
of the PC, and the result is stored in register Z. The updated value is moved from register
Z back into the PC during step 2, while waiting for the memory to respond. In step 3, the
ud
word fetched from the memory is loaded into the IR.

Steps 1 through 3 constitute the instruction fetch phase, which is the same for all
st
instructions. The instruction decoding circuit interprets the contents of the IR at the
ity
beginning of step 4. This enables the control circuitry to activate the control signals for
steps 4 through 7, which constitute the execution phase. The contents of register R3 are
.c
transferred to the MAR in step 4, and a memory read operation is initiated.

Then the contents of Rl are transferred to register Y in step 5, to prepare for the
w
addition operation. When the Read operation is completed, the memory operand is
w
available in register MDR, and the addition operation is performed in step 6. The contents
of MDR are gated to the bus, and thus also to the B input of the ALU, and register Y is
w
selected as the second input to the ALU by choosing Select Y. The sum is stored in
register Z, then transferred to Rl in step 7. The End signal causes a new instruction fetch
cycle to begin by returning to step 1.

This discussion accounts for all control signals in Figure 7.6 except Y in step 2.
There is no need to copy the updated contents of PC into register Y when executing the
Add instruction. But, in Branch instructions the updated value of the PC is needed to
compute the Branch target address. To speed up the execution of Branch instructions, this
value is copied into register Y in step 2. Since step 2 is part of the fetch phase, the same
om
action will be performed for all instructions. This does not cause any harm because
register Y is not used for any other purpose at that time.
c
p.
Branch Instructions:
ou
A branch instruction replaces the contents of the PC with the branch target
address. This address is usually obtained by adding an offset X, which is given in the
gr
branch instruction, to the updated value of the PC. Listing in figure 8 below gives a
control sequence that implements an unconditional branch instruction. Processing starts,
ts
as usual, with the fetch phase. This phase ends when the instruction is loaded into the IR
en
in step 3. The offset value is extracted from the IR by the instruction decoding circuit,
which will also perform sign extension if required. Since the value of the updated PC is
ud
already available in register Y, the offset X is gated onto the bus in step 4, and an
addition operation is performed. The result, which is the branch target address, is loaded
into the PC in step 5.
st
The offset X used in a branch instruction is usually the difference between the branch
ity
target address and the address immediately following the branch instruction.
.c
w
w
w
]
Fig 8

For example, if the branch instruction is at location 2000 and if the branch target
address is 2050, the value of X must be 46. The reason for this can be readily appreciated
from the control sequence in Figure 7. The PC is incremented during the fetch phase,
before knowing the type of instruction being executed. Thus, when the branch address is
om
computed in step 4, the PC value used is the updated value, which points to the
instruction following the branch instruction in the memory.
c
Consider now a conditional branch. In this case, we need to check the status of the
p.
condition codes before loading a new value into the PC. For example, for a Branch-on-
negative (Branch<0) instruction, step 4 is replaced with
ou
Offset-field-of-IRout Add, Zin, If N = 0 then End
gr
Thus, if N = 0 the processor returns to step 1 immediately after step 4. If N = 1, step 5 is
ts
performed to load a new value into the PC, thus performing the branch operation.
en
7.3 MULTIPLE-BUS ORGANIZATION:

ud
The resulting control sequences shown are quite long because only one data item
can be transferred over the bus in a clock cycle. To reduce the number of steps needed,
st
most commercial processors provide multiple internal paths that enable several transfers
ity
to take place in parallel.

Figure 7 depicts a three-bus structure used to connect the registers and the ALU of a
.c
processor. All general-purpose registers are combined into a single block called the
register file. In VLSI technology, the most efficient way to implement a number of
w
registers is in the form of an array of memory cells similar to those used in the imple-
w
mentation of random-access memories (RAMs) described in Chapter 5. The register file

in Figure 9 is said to have three ports. There are two outputs, allowing the contents of two
w
different registers to be accessed simultaneously and have their contents placed on buses
A and B. The third port allows the data on bus C to be loaded into a third register during
the same clock cycle.

Buses A and B are used to transfer the source operands to the A and B inputs of
the ALU, where an arithmetic or logic operation may be performed. The result is
transferred to the destination over bus C. If needed, the ALU may simply pass one of its
two input operands unmodified to bus C. We will call the ALU control signals for such
an operation R=A or R=B. The three-bus arrangement obviates the need for registers Y
om
and Z in Figure 2.
A second feature in Figure 9 is the introduction of the Incremental unit, which is
c
used to increment the PC by 4.. The source for the constant 4 at the ALU input
p.
multiplexer is still useful. It can be used to increment other addresses, such as the
memory addresses in Load Multiple and Store Multiple instructions.
ou
gr
ts
en
ud
st
ity
.c
w
Fig 9
w
Consider the three-operand instruction

Add R4,R5,R6
w

om
Fig 10
c
p.
The control sequence for executing this instruction is given in Figure 10. In step
1, the contents of the PC are passed through the ALU, using the R=B control signal, and
ou
loaded into the MAR to start a memory read operation. At the same time the PC is
incremented by 4. Note that the value loaded into MAR is the original contents of the PC.
gr
The incremented value is loaded into the PC at the end of the clock cycle and will not
affect the contents of MAR. In step 2, the processor waits for MFC and loads the data
ts
received into MDR, then transfers them to IR in step 3. Finally, the execution phase of
en
the instruction requires only one control step to complete, step 4.
By providing more paths for data transfer a significant reduction in the number of
ud
clock cycles needed to execute an instruction is achieved.

st
7.4 HARDWIRED CONTROL:

ity
To execute instructions, the processor must have some means of generating the con-
.c
trol signals needed in the proper sequence. Computer designers use a wide variety of
techniques to solve this problem. The approaches used fall into one of two categories:
w
hardwired control and micro programmed control. We discuss each of these techniques in
w
detail, starting with hardwired control in this section.

Consider the sequence of control signals given in Figure 7. Each step in this
w
sequence is completed in one clock period. A counter may be used to keep track of the
control steps, as shown in Figure 11. Each state, or count, of this counter corresponds to

one control step. The required control signals are determined by the following
information:
1. Contents of the control step counter
2. Contents of the instruction register
3. Contents of the condition code flags
om
4. External input signals, such as MFC and interrupt requests
c
p.
ou
gr
ts
en
ud
st
ity
Fig 11
.c
To gain insight into the structure of the control unit, we start with a simplified
view of the hardware involved. The decoder/encoder block in Figure 11 is a
w
combinational circuit that generates the required control outputs, depending on the state
w
of all its inputs. By separating the decoding and encoding functions, we obtain the more
detailed block diagram in Figure 12. The step decoder provides a separate signal line for
w
each step, or time slot, in the control sequence. Similarly, the output of the instruction
decoder consists of a separate line for each machine instruction. For any instruction
loaded in the IR, one of the output lines INS1 through INSm is set to 1, and all other lines

are set to 0. (For design details of decoders, refer to Appendix A.) The input signals to the
encoder block in Figure 12 are combined to generate the individual control signals Yin,
PCout, Add, End, and so on. An example of how the encoder generates the Zin control
signal for the processor organization in Figure 2 is given in Figure 13. This circuit
implements the logic function
om
Zin=T1+T6 - ADD + T4-BR+---
c
This signal is asserted during time slot Ti for all instructions, during T6 for an Add
p.
instruction, during T4 for an unconditional branch instruction, and so on. The logic
function for Zin is derived from the control sequences in Figures 7 and 8. As another
ou
example, Figure 14 gives a circuit that generates the End control signal from the logic
function
gr
End = T7 ADD + T5 BR + (T5 N + T4 N) BRN +
ts
en
The End signal starts a new instruction fetch cycle by resetting the control step counter to
its starting value. Figure 12 contains another control signal called RUN. When
ud
st
ity
.c
w
w
w
Fig 12

set to 1, RUN causes the counter to be incremented by one at the end of every clock
cycle. When RUN is equal to 0, the counter stops counting. This is needed whenever the
WMFC signal is issued, to cause the processor to wait for the reply from the memory.
c om
p.
ou
Fig 13a
gr
The control hardware shown can be viewed as a state machine that changes from
ts
one state to another in every clock cycle, depending on the contents of the instruction
en
register, the condition codes, and the external inputs. The outputs of the state machine are
the control signals. The sequence of operations carried out by this machine is determined
ud
by the wiring of the logic elements, hence the name "hardwired." A controller that uses
this approach can operate at high speed. However, it has little flexibility, and the
complexity of the instruction set it can implement is limited.
st
ity
.c
w
w
w
Fig 13b

7.5 MICROPROGRAMMED CONTROL:
ALU is the heart of any computing system, while Control unit is its brain. The design
of a control unit is not unique; it varies from designer to designer. Some of the
om
commonly used control logic design methods are;
Sequence Reg & Decoder method
Hard-wired control method
c
PLA control method
p.
Micro-program control method
ou
gr
ts
en
ud
st
The control signals required inside the processor can be generated using a control
step counter and a decoder/ encoder circuit. Now we discuss an alternative scheme, called
ity
micro programmed control, in which control signals are generated by a program similar
to machine language programs.
.c
w
w
w

c om
Fig 15
p.
First, we introduce some common terms. A control word (CW) is a word whose
ou
individual bits represent the various control signals in Figure 12. Each of the control steps
in the control sequence of an instruction defines a unique combination of Is and Os in the
gr
CW. The CWs corresponding to the 7 steps of Figure 6 are shown in Figure 15. We have
assumed that Select Y is represented by Select = 0 and Select4 by Select = 1. A sequence
ts
of CWs corresponding to the control sequence of a machine instruction constitutes the
micro routine for that instruction, and the individual control words in this micro routine
en
are referred to as microinstructions.
The micro routines for all instructions in the instruction set of a computer are
ud
stored in a special memory called the control store. The control unit can generate the
control signals for any instruction by sequentially reading the CWs of the corresponding
st
micro routine from the control store. This suggests organizing the control unit as shown
ity
in Figure 16. To read the control words sequentially from the control store, a micro
program counter (PC) is used. Every time a new instruction is loaded into the IR, the
.c
output of the block labeled "starting address generator" is loaded into the PC. The PC
is then automatically incremented by the clock, causing successive microinstructions to
w
be read from the control store. Hence, the control signals are delivered to various parts of
w
the processor in the correct sequence.

One important function of the control unit cannot be implemented by the simple
w
organization in Figure 16. This is the situation that arises when the control unit is
required to check the status of the condition codes or external inputs to choose between
alternative courses of action. In the case of hardwired control, this situation is handled by

including an appropriate logic function, in the encoder circuitry. In micro programmed

control, an alternative approach is to use conditional branch microinstructions. In
addition to the branch address, these microinstructions specify which of the external
inputs, condition codes, or, possibly, bits of the instruction register should be checked as
a condition for branching to take place.
om
The instruction Branch <0 may now be implemented by a micro routine such as that
shown in Figure 17. After loading this instruction into IR, a branch
c
p.
ou
gr
ts
en
Fig 16
ud
st
ity
.c
w
w
w
Fig 17

microinstruction transfers control to the corresponding micro routine, which is assumed

to start at location 25 in the control store. This address is the output of staring address
generator block codes. If this bit is equal to 0, a branch takes place to location 0 to fetch a
new machine instruction. Otherwise, the microinstruction at location 0 to fetch a new
machine instruction. Otherwise the microinstruction at location 27 loads this address into
om
the PC
c
p.
ou
gr
ts
en
Fig 18
ud
To support micro program branching, the organization of the control unit should
st
be modified as shown in Figure 18. The starting address generator block of Figure 16
becomes the starting and branch address generator. This block loads a new address into
ity
the PC when a microinstruction instructs it to do so. To allow implementation of a

conditional branch, inputs to this block consist of the external inputs and condition codes
.c
as well as the contents of the instruction register. In this control unit, the PC is
w
incremented every time a new microinstruction is fetched from the micro program
memory, except in the following situations:
w
1. When a new instruction is loaded into the IR, the PC is loaded with the starting
w
address of the micro routine for that instruction.

2. When a Branch microinstruction is encountered and the branch condition is satis-
fied, the PC is loaded with the branch address.
3. When an End microinstruction is encountered, the PC is loaded with the address

of the first CW in the micro routine for the instruction fetch cycle
Microinstructions
Having described a scheme for sequencing microinstructions, we now take a closer
om
look at the format of individual microinstructions. A straightforward way to structure
Microinstruction is to assign one bit position to each control signal, as in Figure 15.
c
p.
However, this scheme has one serious drawback assigning individual bits to each
control signal results in long microinstructions because the number of required signals is
ou
usually large. Moreover, only a few bits are set to 1 (to be used for active gating) in any
given microinstruction, which means the available bit space is poorly used. Consider
gr
again the simple processor of Figure 2, and assume that it contains only four general-
purpose registers, R0, Rl, R2, and R3. Some of the connections in this processor are
ts
permanently enabled, such as the output of the IR to the decoding circuits and both inputs
en
to the ALU. The remaining connections to various registers require a total of 20 gating
signals. Additional control signals not shown in the figure are also needed, including the
Read, Write, Select, WMFC, and End signals. Finally, we must specify the function to be
ud
performed by the ALU. Let us assume that 16 functions are provided, including Add,
Subtract, AND, and XOR. These functions depend on the particular ALU used and do not
st
necessarily have a one-to-one correspondence with the machine instruction OP codes. In

ity
total, 42 control signals are needed.

If we use the simple encoding scheme described earlier, 42 bits would be needed
.c
in each microinstruction. Fortunately, the length of the microinstructions can be reduced

easily. Most signals are not needed simultaneously, and many signals are mutually
w
exclusive. For example, only one function of the ALU can be activated at a time. The
w
source for a data transfer must be unique because it is not possible to gate the contents of
two different registers onto the bus at the same time. Read and Write signals to the
w
memory cannot be active simultaneously. This suggests that signals can be grouped so
that all mutually exclusive signals are placed in the same group. Thus, at most one micro
operation per group is specified in any microinstruction. Then it is possible to use a

binary coding scheme to represent the signals within a group. For example, four bits
suffice to represent the 16 available functions in the ALU. Register output control signals
can be placed in a group consisting of PCout, MDRout, Zout, Offsetout, R0out Rlout, R2out,
R3out, and TEMPout. Any one of these can be selected by a unique 4-bit code.
Further natural groupings can be made for the remaining signals. Figure 19
om
shows an example of a partial format for the microinstructions, in which each group
occupies a field large enough to contain the required codes. Most fields must include one
c
inactive code for the case in which no action is required. For example, the all-zero pattern
p.
in Fl indicates that none of the registers that may be specified in this field should have its
contents placed on the bus. An inactive code is not needed in all fields. For example, F4
ou
contains 4 bits that specify one of the 16 operations performed in the ALU. Since no
spare code is included, the ALU is active during the execution of every microinstruction.
gr
However, its activity is monitored by the rest of the machine through register Z, which is
loaded only when the Zin signal is activated.
ts
Grouping control signals into fields requires a little more hardware because
en
decoding circuits must be used to decode the bit patterns of each field into individual
control signals. The cost of this additional hardware is more than offset by the reduced
number of bits in each microinstruction, which results in a smaller control store. In Figure
ud
19, only 20 bits are needed to store the patterns for the 42 signals.
So far we have considered grouping and encoding only mutually exclusive
st
control signals. We can extend this idea by enumerating the patterns of required signals
ity
in all possible microinstructions. Each meaningful combination of active control signals

can
.c
w
w
w

c om
p.
ou
gr
Fig 19
ts
then be assigned a distinct code that represents the microinstruction. Such full
en
encoding is likely to further reduce the length of micro words but also to increase the
complexity of the required decoder circuits.
ud
Highly encoded schemes that use compact codes to specify only a small number of
control functions in each microinstruction are referred to as a vertical organization. On
st
the other hand, the minimally encoded scheme of Figure 15, in which many resources can
be controlled with a single microinstruction, is called a horizontal organization. The
ity
horizontal approach is useful when a higher operating speed is desired and when the
machine structure allows parallel use of resources. The vertical approach results in
.c
considerably slower operating speeds because more microinstructions are needed to

w
perform the desired control functions. Although fewer bits are required for each
microinstruction, this does not imply that the total number of bits in the control store is
w
smaller. The significant factor is that less hardware is needed to handle the execution of
w
microinstructions. Horizontal and vertical organizations represent the two organizational

extremes in micro programmed control. Many intermediate schemes are also possible, in
which the degree of encoding is a design parameter. The layout in Figure 19 is a
horizontal organization because it groups only mutually exclusive micro operations in the

same fields. As a result, it does not limit in any way the processor's ability to perform
various micro operations in parallel.
Although we have considered only a subset of all the possible control signals, this
subset is representative of actual requirements. We have omitted some details that are not
essential for understanding the principles of operation.
c om
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
w

UNIT - 8
Multicores, Multiprocessors, and Clusters: Performance, The Power
Wall, The Switch from Uniprocessors to Multiprocessors, Amdahls Law,
om
Shared Memory Multiprocessors, Clusters and other Message Passing
Multiprocessors, Hardware Multithreading, SISD, IMD, SIMD, SPMD, and
c
Vector.
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
w

UNIT - 8
MULTICORES, MULTIPROCESSORS, AND CLUSTERS
8.1 PERFORMANCE:
om
Computer performance is characterized by the amount of useful work accomplished by a
computer system compared to the time and resources used. Depending on the context,
c
good computer performance may involve one or more of the following:
p.
Short response time for a given piece of work
ou
High throughput (rate of processing work)
Low utilization of computing resource(s)
gr
High availability of the computing system or application
Fast (or highly compact) data compression and decompression
ts
High bandwidth / short data transmission time
en
Technical performance metrics

ud
There are a wide variety of technical performance metrics that indirectly affect overall
computer performance. Because there are too many programs to test a CPU's speed on all
st
of them, benchmarks were developed. The most famous benchmarks are the SPECint
ity
and SPECfpbenchmarks developed by Standard Performance Evaluation Corporation and

the ConsumerMark benchmark developed by the Embedded Microprocessor Benchmark
.c
Consortium EEMBC.
w
Some important measurements include:

w
Instructions per second Most consumers pick a computer architecture

w
(normally Intel IA32 architecture) to be able to run a large base of pre-existing, pre-
compiled software. Being relatively uninformed on computer benchmarks, some of them
pick a particular CPU based on operating frequency (see megahertz myth).

FLOPS The number of floating-point operations per second is often important in

selecting computers for scientific computations.
Performance per watt System designers building parallel computers, such as Google,
pick CPUs based on their speed per watt of power, because the cost of powering the CPU
outweighs the cost of the CPU itself. [2][3]
om
Some system designers building parallel computers pick CPUs based on the speed per
dollar.
c
System designers building real-time computing systems want to guarantee worst-case
p.
response. That is easier to do when the CPU has low interrupt latency and when it has
deterministic response. (DSP[disambiguation needed][clarification needed])
ou
Computer programmers who program directly in assembly language want a CPU to
support a full-featured instruction set.
gr
Low power For systems with limited power sources (e.g. solar, batteries, human
power). ts
Small size or low weight - for portable embedded systems, systems for spacecraft.
en
Environmental impact Minimizing environmental impact of computers during
manufacturing and recycling as well as during use. Reducing waste, reducing hazardous
ud
materials. (see Green computing).

Giga-updates per second - a measure of how frequently the RAM can be updated
st
Occasionally a CPU designer can find a way to make a CPU with better overall
performance by improving one of these technical performance metrics without sacrificing
ity
any other (relevant) technical performance metricfor example, building the CPU out of
better, faster transistors. However, sometimes pushing one technical performance metric
.c
to an extreme leads to a CPU with worse overall performance, because other important
w
technical performance metrics were sacrificed to get one impressive-looking number

for example, the megahertz myth.
w
w

8.2 The Power Wall:

A multi-core processor is a single computing component with two or more
independent actual central processing units (called "cores"), which are the units that read
and execute program instructions.
om
[1] The instructions are ordinary CPU instructions such as add, move data, and branch,
but the multiple cores can run multiple instructions at the same time, increasing overall
speed for programs amenable to parallel computing.
c
[2] Manufacturers typically integrate the cores onto a single integrated circuit die (known
p.
as a chip multiprocessor or CMP), or onto multiple dies in a single chip package.
ou
Processors were originally developed with only one core. A dual-core
processor has two cores (e.g. AMD Phenom II X2, Intel Core Duo), a quad-core
gr
processor contains four cores (e.g. AMD Phenom II X4, Intel's quad-core processors,
see i3, i5, and i7 atIntel Core), a hexa-core processor contains six cores (e.g. AMD
ts
Phenom II X6, Intel Core i7 Extreme Edition 980X), an octa-core processor contains
en
eight cores (e.g. Intel Xeon E7-2820, AMD FX-8150). A multi-core processor
implements multiprocessing in a single physical package. Designers may couple cores in
ud
a multi-core device tightly or loosely. For example, cores may or may not share caches,
and they may implement message passing or shared memory inter-core communication
methods. Common network topologies to interconnect cores include bus, ring, two-
st
dimensional mesh, and crossbar. Homogeneous multi-core systems include only identical
ity
cores, heterogeneous multi-core systems have cores that are not identical. Just as with
single-processor systems, cores in multi-core systems may implement architectures such
.c
as superscalar, VLIW, vector processing, SIMD, or multithreading.

w
Multi-core processors are widely used across many application domains

w
including general-purpose, embedded, network, digital signal processing (DSP),

and graphics.
w
The improvement in performance gained by the use of a multi-core processor

depends very much on the software algorithms used and their implementation. In
particular, possible gains are limited by the fraction of the software that can be run in

parallel simultaneously on multiple cores; this effect is described by Amdahl's law. In the
best case, so-called embarrassingly parallel problems may realize speedup factors near
the number of cores, or even more if the problem is split up enough to fit within each
core's cache(s), avoiding use of much slower main system memory. Most applications,
however, are not accelerated so much unless programmers invest a prohibitive amount of
om
effort in re-factoring the whole problem.
[3] The parallelization of software is a significant ongoing topic of research.
c
p.
ou
8.3 The Switch from Uniprocessors to Multiprocessors:
The steps are:
gr
In Control Panel, click System, choose the Hardware tab, then click the Device
Manager button.
ts
en
Select the Computer node and expand it.
Double-click the object listed there (on my system, it is called Standard PC), then
ud
choose the Driver tab. Click the Update Driver button.

st
On the Upgrade Device Driver Wizard, click the Next button, then select Display
a known list of drivers for this device so that I can choose a specific driver. Click
ity
the Next button.

.c
On the Select Device Driver page, select Show all hardware of this device class.
w
Select the HAL that matches your new configuration, multiprocessor or

w
uniprocessor. Click the Next button. Check that the wizard is showing the
w
configuration you want, then click the Next button to complete the wizard.

8.4 Amdahls Law:

Amdahl's law, also known as Amdahl's argument is named after computer
architect Gene Amdahl, and is used to find the maximum expected improvement to an
om
overall system when only part of the system is improved. It is often used in parallel
computing to predict the theoretical maximum speedup using multiple processors. It was
presented at the AFIPS Spring Joint Computer Conference in 1967.
c
The speedup of a program using multiple processors in parallel computing is
p.
limited by the time needed for the sequential fraction of the program. For example, if a
program needs 20 hours using a single processor core, and a particular portion of 1 hour
ou
cannot be parallelized, while the remaining promising portion of 19 hours (95%) can be
parallelized, then regardless of how many processors we devote to a parallelized
gr
execution of this program, the minimum execution time cannot be less than that critical 1
hour. Hence the speedup is limited up to 20, as the diagram illustrates.
ts
Amdahl's law is a model for the relationship between the expected speedup of
en
parallelized implementations of an algorithm relative to the serial algorithm, under the
assumption that the problem size remains the same when parallelized. For example, if for
ud
a given problem size a parallelized implementation of an algorithm can run 12% of the
algorithm's operations arbitrarily quickly (while the remaining 88% of the operations are
st
not parallelizable), Amdahl's law states that the maximum speedup of the parallelized
version is 1/(1 0.12) = 1.136 times as fast as the non-parallelized implementation.
ity
More technically, the law is concerned with the speedup achievable from an
.c
improvement to a computation that affects a proportion P of that computation where the

improvement has a speedup of S. (For example, if 30% of the computation may be the
w
subject of a speed up, P will be 0.3; if the improvement makes the portion affected twice
w
as fast, S will be 2.) Amdahl's law states that the overall speedup of applying the
improvement will be:
w

To see how this formula was derived, assume that the running time of the old
computation was 1, for some unit of time. The running time of the new computation will
be the length of time the unimproved fraction takes (which is 1 - P), plus the length of
time the improved fraction takes. The length of time for the improved part of the
computation is the length of the improved part's former running time divided by the
om
speedup, making the length of time of the improved part P/S. The final speedup is
computed by dividing the old running time by the new running time, which is what the
c
above formula does.
p.
Here's another example. We are given a sequential task which is split into four
ou
consecutive parts: P1, P2, P3 and P4 with the percentages of runtime being 11%, 18%,
23% and 48% respectively. Then we are told that P1 is not sped up, so S1 = 1, while P2 is
sped up 5, P3 is sped up 20, and P4 is sped up 1.6. By using the formula P1/S1 +
gr
P2/S2 + P3/S3 + P4/S4, we find the new sequential running time is: or a little less
ts
than 12 the original running time. Using the formula (P1/S1 + P2/S2 + P3/S3 +
P4/S4)-1, the overall speed boost is 1 / 0.4575 = 2.186, or a little more than double the
en
original speed. Notice how the 20 and 5 speedup don't have much effect on the overall
speed when P1 (11%) is not sped up, and P4 (48%) is sped up only 1.6 times.
ud
st
8.5 Shared Memory Multiprocessors shared memory multiprocessor:

Shared memory multiprocessors have multiple CPUs all of which share the same
ity
address space. This means there is only one memory accessed by all CPUs on an equal
basis. Shared memory systems can be either SIMD or MIMD. Single-CPU
.c
vectorprocessors can be regarded as an example of the former, while the multi-CPU

w
models of these machines are examples of the latter. The main problem with shared
memory systems is the connection of the CPUs to each other and to the memory. Various
w
interconnection alternatives have been used, including crossbar, $\Omega$-network, and

w
central databus. The IBM SP2, the Meiko CS-2, and the Cenju-3 use the $\Omega$-
network.\cite{van-der-Steen-overview}

8.6 Clusters and other Message Passing Multiprocessors:
What is MPI?
1.MPI stands for Message Passing Interface and its standard is set by the Message
om
Passing Interface Forum
2.It is a library of subroutines/functions, NOT a language
3.MPI subroutines are callable from Fortran and C
c
4.Programmer writes Fortran/C code with appropriate MPI library calls, compiles
p.
with Fortran/C compiler, then links with Message Passing library
5. For large problems that demand better turn-around time (and access to more
ou
memory)
6.For Fortran dusty deck, often it would be very time-consuming to rewrite code
gr
to take advantage of parallelism. Even in the case of SMP, as are the SGI
PowerChallengeArray and Origin2000, automatic parallelizer might not be able to
ts
detect parallelism.
en
7.For distributed memory machines, such as a cluster of Unix work stations or a
cluster of NT/Linux PCs.
ud
8.Maximize portability; works on distributed and shared memory architectures.
Preliminaries of MPI Message Passing:

st
In a user code, wherever MPI library calls occur, the following header file must be
ity
included:
#include mpi.h for C code or
.c
include mpif.h for Fortran code

These files contain definitions of constants, prototypes, etc. which are neccessary
w
to compile a program that contains MPI library calls

w
MPI is initiated by a call to MPI_Init. This MPI routine must be called before any
other MPI routines and it must only be called once in the program.
w
MPI processing ends with a call to MPI_Finalize.

Essentially the only difference between MPI subroutines (for Fortran programs) and
MPI functions (for C programs) is the error reporting flag. In fortran, it is returned as

the last member of the subroutines argument list. In C, the integer error flag is
returned through the function value. Consequently, MPI fortran routines always
contain one additional variable in the argument list than the C counterpart.
Cs MPI function names start with MPI_ and followed by a character string with
the leading character in upper case letter while the rest in lower case letters. Fortran
om
subroutines bear the same names but are case-insensitive.
c
8.7 Hardware Multithreading:
p.
Multithreading computer central processing units have hardware support to
efficiently execute multiple threads. These are distinguished
ou
from multiprocessing systems (such as multi-core systems) in that the threads have to
share the resources of a single core: the computing units, the CPU caches and
gr
the translation lookaside buffer(TLB). Where multiprocessing systems include multiple
complete processing units, multithreading aims to increase utilization of a single core by
ts
using thread-level as well as instruction-level parallelism. As the two techniques are
en
complementary, they are sometimes combined in systems with multiple multithreading
CPUs and in CPUs with multiple multithreading cores.
ud
Overview
The multithreading paradigm has become more popular as efforts to further

st
exploit instruction level parallelism have stalled since the late-1990s. This allowed the
ity
concept of throughput computing to re-emerge to prominence from the more specialized

field of transaction processing:
.c
Even though it is very difficult to further speed up a single thread or single program,
w
most computer systems are actually multi-tasking among multiple threads or programs.
Techniques that would allow speedup of the overall system throughput of all tasks
w
would be a meaningful performance gain.

w
The two major techniques for throughput computing are multiprocessing and
multithreading.

Some advantages include:
If a thread gets a lot of cache misses, the other thread(s) can continue, taking advantage
of the unused computing resources, which thus can lead to faster overall execution, as
these resources would have been idle if only a single thread was executed.
om
If a thread cannot use all the computing resources of the CPU (because instructions
depend on each other's result), running another thread can avoid leaving these idle.
If several threads work on the same set of data, they can actually share their cache,
c
leading to better cache usage or synchronization on its values.
p.
ou
Some criticisms of multithreading include:
gr
Multiple threads can interfere with each other when sharing hardware resources such as
caches or translation lookaside buffers (TLBs).
ts
Execution times of a single thread are not improved but can be degraded, even when
en
only one thread is executing. This is due to slower frequencies and/or additional pipeline
stages that are necessary to accommodate thread-switching hardware.
ud
Hardware support for multithreading is more visible to software, thus requiring more
changes to both application programs and operating systems than multiprocessing.
The mileage thus varies; Intel claims up to 30 percent improvement with
st
its HyperThreading technology, while a synthetic program just performing a loop of non-
ity
optimized dependent floating-point operations actually gains a 100 percent speed

improvement when run in parallel. On the other hand, hand-tuned assembly
.c
languageprograms using MMX or Altivec extensions and performing data pre-fetches (as
a good video encoder might), do not suffer from cache misses or idle computing
w
resources. Such programs therefore do not benefit from hardware multithreading and can
w
indeed see degraded performance due to contention for shared resources.

w
Hardware techniques used to support multithreading often parallel the software

techniques used for computer multitasking of computer programs.

Thread scheduling is also a major problem in multithreading.

Concept
The simplest type of multi-threading occurs when one thread runs until it is blocked
by an event that normally would create a long latency stall. Such a stall might be a cache-
miss that has to access off-chip memory, which might take hundreds of CPU cycles for
om
the data to return. Instead of waiting for the stall to resolve, a threaded processor would
switch execution to another thread that was ready to run. Only when the data for the
c
previous thread had arrived, would the previous thread be placed back on the list
p.
of ready-to-run threads.
ou
For example:
1.Cycle i : instruction j from thread A is issued
gr
2.Cycle i+1: instruction j+1 from thread A is issued
3.Cycle i+2: instruction j+2 from thread A is issued, load instruction which misses in all
caches
ts
en
4.Cycle i+3: thread scheduler invoked, switches to thread B
5.Cycle i+4: instruction k from thread B is issued
ud
6.Cycle i+5: instruction k+1 from thread B is issued

Conceptually, it is similar to cooperative multi-tasking used in real-time operating
st
systems in which tasks voluntarily give up execution time when they need to wait upon
some type of the event.
ity
Terminology:
.c
This type of multi threading is known as Block or Cooperative or Coarse-

grained multithreading.
w
Hardware cost:
w
The goal of multi-threading hardware support is to allow quick switching between

w
a blocked thread and another thread ready to run. To achieve this goal, the hardware cost
is to replicate the program visible registers as well as some processor control registers
(such as the program counter). Switching from one thread to another thread means the
hardware switches from using one register set to another.

Such additional hardware has these benefits:
The thread switch can be done in one CPU cycle.

It appears to each thread that it is executing alone and not sharing any hardware
resources with any other threads[dubious discuss]. This minimizes the amount of
om
software changes needed within the application as well as the operating system to support
multithreading.
In order to switch efficiently between active threads, each active thread needs to have
c
its own register set. For example, to quickly switch between two threads, the register
p.
hardware needs to be instantiated twice.
ou
Examples
Many families of microcontrollers and embedded processors have multiple register
gr
banks to allow quick context switching for interrupts. Such schemes can be considered a
ts
type of block multithreading among the user program thread and the interrupt threads
en
1.Cycle i+1: an instruction from thread B is issued
2.Cycle i+2: an instruction from thread C is issued
ud
The purpose of interleaved multithreading is to remove all data dependency stalls from
the execution pipeline. Since one thread is relatively independent from other threads,
st
there's less chance of one instruction in one pipe stage needing an output from an older
instruction in the pipeline. Conceptually, it is similar to pre-emptive multi-tasking used
ity
in operating systems. One can make the analogy that the time-slice given to each active
thread is one CPU cycle.
.c
Hardware costs:
w
In addition to the hardware costs discussed for interleaved multithreading, SMT

w
has the additional cost of each pipeline stage tracking the Thread ID of each instruction
being processed. Again, shared resources such as caches and TLBs have to be sized for
w
the large number of active threads being processed.

Examples
DEC (later Compaq) EV8 (not completed)
Intel Hyper-Threading
IBM POWER5
Sun Microsystems UltraSPARC T2
om
MIPS MT
CRAY XMT
c
p.
8.8 SISD, IMD, SIMD, SPMD, and Vector:
Single instruction, multiple data (SIMD), is a class of parallel computers in Flynn's
ou
taxonomy. It describes computers with multiple processing elements that perform the
same operation on multiple data points simultaneously. Thus, such machines exploit data
gr
level parallelism. SIMD is particularly applicable to common tasks like adjusting the
contrast in a digital image or
ts adjusting the volume of digital audio. Most
modern CPU designs include SIMD instructions in order to improve the performance
en
of multimedia use.
ud
In computing, SPMD (single program, multiple data) is a technique employed to

achieve parallelism; it is a subcategory of MIMD. Tasks are split up and run
simultaneously on multiple processors with different input in order to obtain results
st
faster. SPMD is the most common style of parallel programming.[1] It is also a

ity
prerequisite for research concepts such as active messages and distributed shared
memory.
.c
SPMD vs SIMD
w
In SPMD, multiple autonomous processors simultaneously execute the same program at

w
independent points, rather than in the lockstep that SIMD imposes on different data. With
w
SPMD, tasks can be executed on general purpose CPUs; SIMD requires vector
processors to manipulate data streams. Note that the two are not mutually exclusive.

Distributed memory
SPMD usually refers to message passing programming on distributed
memory computer architectures. A distributed memory computer consists of a collection
of independent computers, called nodes. Each node starts its own program and
communicates with other nodes by sending and receiving messages, calling send/receive
om
routines for that purpose. Barrier synchronization may also be implemented by messages.
The messages can be sent by a number of communication mechanisms, such
c
as TCP/IP over Ethernet, or specialized high-speed interconnects such
p.
as Myrinet and Supercomputer Interconnect. Serial sections of the program are
implemented by identical computation on all nodes rather than computing the result on
ou
one node and sending it to the others. Nowadays, the programmer is isolated from the
details of the message passing by standard interfaces, such as PVM and MPI.
gr
Distributed memory is the programming style used on parallel supercomputers from
ts
homegrown Beowulf clusters to the largest clusters on the Teragrid.
en
Shared memory:
On a shared memory machine (a computer with several CPUs that access the same
ud
memory space), messages can be sent by depositing their contents in a shared memory
area. This is often the most efficient way to program shared memory computers with
st
large number of processors, especially on NUMA machines, where memory is local to

processors and accessing memory of another processor takes longer. SPMD on a shared
ity
memory machine is usually implemented by standard (heavyweight) processes.

.c
Unlike SPMD, shared memory multiprocessing, also called symmetric

multiprocessing or SMP, presents the programmer with a common memory space and the
w
possibility to parallelize execution by having the program take different paths on different
w
processors. The program starts executing on one processor and the execution splits in a
parallel region, which is started when parallel directives are encountered. In a parallel
w
region, the processors execute a single program on different data. A typical example is
the parallel DO loop, where different processors work on separate parts of the arrays
involved in the loop. At the end of the loop, execution is synchronized, only one

processor continues, and the others wait. The current standard interface for shared
memory multiprocessing is OpenMP. It usually implemented by lightweight processes,
called threads.
Combination of levels of parallelism
om
Current computers allow exploiting of many parallel modes at the same time for
maximum combined effect. A distributed memory program using MPI may run on a
collection of nodes. Each node may be a shared memory computer and execute in parallel
c
on multiple CPUs using OpenMP. Within each CPU, SIMD vector instructions (usually
p.
generated automatically by the compiler) and superscalar instruction execution (usually
ou
handled transparently by the CPU itself), such as pipeliningand the use of multiple
parallel functional units, are used for maximum single CPU speed.
In computing, MIMD (multiple instruction, multiple data) is a technique employed
gr
to achieve parallelism. Machines using MIMD have a number of processors that
ts
function asynchronously and independently. At any time, different processors may be
executing different instructions on different pieces of data. MIMD architectures may be
en
used in a number of application areas such as computer-aided design/computer-aided
manufacturing, simulation, modeling, and as communication switches. MIMD machines
ud
can be of either shared memory or distributed memory categories. These classifications

are based on how MIMD processors access memory. Shared memory machines may be
st
of the bus-based, extended, or hierarchical type. Distributed memory machines may

have hypercube or mesh interconnection schemes.
ity
Bus-based
.c
MIMD machines with shared memory have processors which share a common,
w
central memory. In the simplest form, all processors are attached to a bus which connects
them to memory.
w
w

Computer organization 10CS46
Assignment Questions
UNIT - 1
Basic Structure of Computers
om
1. Explain different functional units of a computer
2. With a neat diagram explain different processor registers.
3. What is a bus..? explain single bus and multiple bus structure used to inter-connect
c
functional units in computer system
p.
4. Explain how the performance of a computer can be measured. What are the measures to
improve the performance of computers..?*
ou
5. Explain the important technological features and devices that characterized each
generation
gr
6. Explain (i)byte addressability (ii)big-endian assignment (iii)little endian assignment
ts
en
UNIT - 2
Machine Instructions and Programs contd...
ud
1. What is an addressing mode? Explain different generic addressing modes with an

example for each.
st
2. What are assembler directives..?explain any two directives

ity
3. For a simple example of I/O operations involving a keyboard and a display device, write
a assembly language program that reads one line from the keyboard, stores it in memory
.c
buffer and echoes it back to the display.

4. Define subroutine .explain subroutine linkage using a link register.
w
5. Explain with an example, usage of stacks in a nested subroutine calls.

w
6. Explain how stacks and queues are organized.

w
Dept of CSE ,SJBIT Page 1

UNIT - 3
Input/output Organization
1. Explain the followings:

(i)Interrupt concepts (ii) interrupt hardware.
om
2. With neat block diagram , explain any two methods of handling multiple I/O devices
3. Define exceptions. Explain kinds of exceptions.*
c
4. What is the necessity of DMA controller? Explain the methods of bus arbitration.
p.
5. Show the possible register configurations in a DMA interface, explain direct memory
access(DMA)*
ou
6. What is the necessity of bus arbitration? Explain the different methods of bus arbitration.
UNIT - 4
gr
Input/output Organization contd...
ts
1. Compare programmed I/O. interrupt driven I/O and DMA based I/O
en
2. Compare serial and parallel interfaces for efficiency and complexity with example
3. What are the features of SCSI bus? Write a note on arbitration and selection on SCSI bus
4. Describe the split bus operation. How can it be connected to two fast devices and one slow
ud
device?
st
UNIT - 5
Memory System
ity
1. How read and write operation takes place in 1k *1 memory chip ?explain
.c
2. With the block diagram explain the operation of a 16-megabit DRAM chip configured as
w
2M*8.
3. Mention any two differences b/w static and dynamic RAMs . explain the internal
w
organization of a memory chip consisting of 16 words of 8 bit each.*

w
4. Which are the various factors to be considered in the choice of a memory chip ? explain
5. Give the organization of a 2M*32 memory module using 512k*8 static memory chips
6. Discuss ,the different types of RAMs bring out their salient features

UNIT - 6
Arithmetic
1. Write a note on flash memory.

2. What is secondary storage? Explain in brief magnetic hard disk.
om
3. Explain about optical disks and its applications; explain working principle of CD
technology.
4. A half adder is a combinational logic circuit that has two inputs, x and y and two outputs,
c
s and c , that are the sum and carry-out respectively, resulting from the binary addition of
p.
x and y:
ou
(i). design a half adder as a two level AND OR circuit.
(ii). Show how to implement a full- adder using two half-adder and external
gr
Logic gates are necessary
ts
5. Explain how a 16-bit carry look ahead adder can be built from 4 bit adder.
en
6. How do you design FAST ADDERS? Explain a 4 bit carry look ahead adder.
ud
UNIT - 7
Basic Processing Unit
st
1. Explain hardwired control unit with diagram.

ity
2. Write and explain control sequences for execution of the instruction SUB R1, (R4).
3. Draw the block diagram of a processor and identify the units

.c
4. Explain three bus architecture with diagram

w
UNIT - 8
w
Multicores, Multiprocessors, and Clusters

w
1. How TLB can be used in implementing virtual memory? Explain with Block diagram
2. Explain 4 bit carry-look ahead adder and use it to build a 12 bit carry-look ahead adder
3. Perform 56-78 using 1s complement and 2s complement methods.

4. Explain the process of fetching a word from memory, with timing diagram
5. Explain about optical disks and its applications; explain working principle of CD
technology
om
6. What is the necessity of bus arbitration? Explain the different methods of bus arbitration.
c
p.
ou
gr
ts
en
ud
st
ity
.c
w
w
w

Computer Organization 10CS46
VTU QUESTION BANK
UNIT-I: BASIC STRUCTURE OF COM PUTERS
om
1. With a neat diagram explain different processor registers June 2012(8M)
generation of computers Dec 2013, Jan 2012(10M)
c
3. Discuss two ways in which byte addresses are assigned June 2012(10M)
p.
4. Explain the different functional units of a computer with a neat block diagram
ou
Dec 2013(10M)
5. Write the basic performance equation. explain the role of each of the parameters in
the equation on the performance of the computer. June 2013(12M)
gr
6. With a neat diagram ,discuss the basic operational concepts of a computer
July 2013(10M)
ts
7. List the different systems used to repres ent a signed number Jan 2014(10M)
en
8. Explain with necessary block diagram the basic functional unit of a computer.
Jan 2014, Jan 2012( 1 0 M )
ud
UNIT-II- MACHINE INSTRUCTIONS AND PROGRAMS CONTD.:

st
1. What is an addressing m ode .Explain different generic addressing modes with an

ity
example for each June 2012 /July 2013 /Jan 14
2. Define subroutine. Explain subroutine linkage using a link register Dec 2013
.c
3. Explain the shift and rotate operation with examples June 2012
w
4. What is the need for an addressing mode? Explain the following addressing modes
w
with examples: immediate, direct, indirect, index, relative Dec 2013 /July 2013
w
5. a)What is subroutine linkage? how are parameters passed to subroutines June 2012
b.What is stack frame? Explain
6. Discuss briefly encoding of machine instructions. July 2013
DEPT OF CSE,SJBIT Page 1

7. Big-Endian and Little-Endian assignment,explain with necessary figure. Jan 2014
8. List the name, assembler syntax and addressing functions for the dufferent addressing
modes. Jan 2014
9. Explain logical and arithmetic shift instructions with an example. Jan 2014
om
UNIT-III- INPUT/OUTPUT ORGANIZATION
c
1. A. Explain the following (i) Interrupt concepts (ii) Interrupt hardware
p.
B. Define exceptions. Explain two kinds of exceptions
ou
C. What is the necessity of bus arbitration .Explain the different methods of bus
arbitration. July 2013
gr
D. Define the following(i) cycle stealing (ii) burst mode June 2013
ts
2. A. discuss the different schemes available to disable and enable the interrupts
en
B. how are simultaneous interrupt from more than one devices handled? Ju l y 2013
C. What does the term cycle stealing mean?

ud
D. Write a short note on any 1 bus arbitration scheme.(10M)Dec 2013

st
3. A .define and explain briefly the following

ity
i) Interrupt
.c
ii) Vectored interrupt
iii) Interrupt nesting

w
B. Explain in with the help of a diagram the working of daisy chain with multiple
w
priority levels and multiple devices in each level June 2014

w
4. A. In modern computers, why interrupts are required ? Support your claim with suitable
example

B. In the interrupt mechanism, how the simultaneous arrivals of interrupts from various
(multiple) devices (I/O) are handled? July 2014
C. With neat sketches, Ex plain the various approaches to bus arbitration. Dec 2014
om
5. A. Explain with a diagram, how interrupt request from several I/O devices can be
communicated to a processor through a single INTR line
c
B. How can the processor obtain the starting address of different interrupt service
p.
routines usingvectored interrupts?
ou
C. Why is bus arbitration required? Explain with block diagram bus arbitration using
daisy chain. June 20124
gr
ts
6.a) Draw t h e a r r angeme n t of a sin g l e bus str u c ture and brie f abou t memory mapped
en
I/O. Jan 2014
b) Exp l a in i)i n t erru p t enabling

ud
ii)edge tr i g g ering with respect interr u p ts.
c) Draw t h e arrang e ments for bus arbitr a t ion using a d a i sy c h ain a n d ex p l ai n in

st
brief.
ity
UNIT-IV- INPUT/OUTPUT ORGANIZATION CONTD.

.c
1. A. Explain the important functions input/output interface July 2013

w
B. With a neat block diagram explain a general 8-bit parallel interface circuit
w
C. Explain the following w.r.t. USB(i) USB architecture (ii) USB addressing (iii) USB
w
protocols June2013 / J u l y 2 01 3
2. A.Draw and explain the block diagram of a typical serial interface. How does it

compare with a parallel interface?
B. explain the main phases involved in SCSI bus operation Dec2013
3. A. in a computer system, PCI bus is used to connect devices to the processor (system
om
bus) bus. Consider a bus transactional which the processor reads four 32-bit words
from the memory. Explain the read operation on the PCI bus between memory and
processor .give signal and timing diagram
c
B. Draw the block diagram of universal bus(USB)structure connected to the host
p.
computer Briefly explain all fields of packets that are used for communication between a
ou
host and a device connected to an USB port.June2014 / J u l y 2 0 1 3
gr
4. A .With a neat sketch, explain the individual input and output interface circuits. Also
elicit their salient feature.
ts
B. In a computer system why a PCI bus is used? With a neat sketch, explain how the read
en
operation is performed along with the role of IRDY#/TRDY# on the PCI bus (10)
Dec2014
ud
5. A. Explain with a block diagram a general 8 bit parallel interface.

st
B. With the help of a data transfer signals explain how a real operation is performed
ity
u sing PCI bus.

.c
C. Explain briefly bus arbitration phase in SCSI bus. June 2012
6. a. Draw the hardware components needed for connecting a keyboard to a process and
w
explain in brief. Jan 2014

w
b. Explain the use of PCI bus in a computer system with necessary figure.
w
c. List the SCSI bus signals with their functionalities.

UNIT-V- MEMORY SYSTEM
1. A. with the block diagram, explain the operations of 16 megabit DRAM configured as
2Mx8 June 2013,JAN 2014.
om
B. Explain different mapping functions used in cache memory
C. What do you mean by memory interleaving? Explain. June 2013
c
p.
2. A. differentiate between SRAM and DRAM Dec 2013
B. Sketch and explain the internal organization of a 2Mx8 dynamic memory chip
ou
C. Explain any 1 cache mapping function. Dec 2013
gr
3. A. Define and explain the following
i) Memory access time

ts ii) memory cycle time
iii) Random access memory(RAM) iv)static memory

en
B. Differentiate the static RAM (SRAM) and dynamic RAM (DRAM) giving four ke y
ud
differences. State the primary usage of SRAM and DRAM in contemporary computer
s ystems June 2014
st
C. Define memory latency and bandwidth in case of burst operation that is used for
transferring a block of data to or from synchronous DRAM memory unit June 2014
ity
.c
4. A draw a diagram and explain the working of a 16-M bit DRAM chip configured as
w
2M x8 Also explain as to how it can be made to work in fast page mode Dec 2014
w
5. A. Draw the organization of a 1Kx1 memory cell and explain its working.
B. Explain the working of a single-transistor dynamic memory cell

w
C. Calculate the average access time experienced by a processor if a cache bit rate is
0.88,miss penalty is 0.015 milliseconds and cache access time is 10 microseconds.
June 2012

6. Ex p l a i n any two ca c h e mapping f u nc t i o n s. J u l y 2 013
7. Ex p l ain t h e t r a n s l a t i o n of virt u a l ad d r e s s t o p hys i ca l a d d r e s s s Ju l y 2 0 1 3
8. a. Draw f o r 1K * 1 me m o ry c h i p w i t h neat figure. J a n 2 0 14
b. S h o w w i t h d i a g r a m t h e memory hie r a rchy w i t h re s p e c t to s p eed , s i ze a n d c o s t .
om
c. W i t h figure explain a b o u t di r e c t ma p p ing c a c he me m o ry.
c
UNIT-VI: ARITHMETIC
p.
ou
1. A. what do you mean by address translation. Explain how TLB is used to implement
virtual memory
gr
B. Mention four major functions of disk controller on disk drive side.
C. Explain how a 16-bit carry look ahead adder can be built from 4-bit adder
ts
Dec 2013
en
2. A. Draw a block diagram and explain how a virtual address from the pro cessor is
translated into a physical address in the main memory
ud
B. write short notes on (i) optical technology used in CD systems (ii) RAID disk
arrays(8M)
st
C. Draw a figure to illustrate and explain a 16-bit carry look ahead adder using 4-bit
adder blocks. Show that the sum and carry are generated in 5 and 8 gate dela ys
ity
respectivel y(6M) June 2013
3 .A. Explain a simple method of translating virtual address of a program into physical
.c
add ress, with the help of a diagram J u l y 2013

w
B. Explain structural organization of moving head magnetic hard disk, with multiple
w
surfaces for storage of data. Explain hw moving head assembly works for reading data
Dec 2013
w
C. Answer the following w.r.t. to magnetic disk, the secondary storage device(6M)
i) Seek time
ii) Latency

iii) Access timeJune 2014
4. Explain in detail the working principle of a magnetic hard diskDec 2014
om
5. A. Show the organization of virtual memory address translation based in fixed-length
pages a nd explains its workin g.
B. How can performance and reality be improved using RAID technology?
c
C. Explain the design of a 4-bit carry-look a head adder.June 2012 /j u l y 2013
p.
6.a ) Explain with figur e the design of a 4 -bi t carry look adder. Jan 2014
ou
b ) W i t h figu r e expla i n cir c u it arrangement s for binary d i v i s ion.
gr
c) IEE E standard f o r floating poin t n u m bers,explain.
ts
UNIT-VII: BASIC PROCESSING UNIT
en
1 .A. Show the multiplication of (+13) and (-6) using multiplier bit pair recording
ud
technique.
B. Differentiate between restoring and non restoring division July 2013

st
C. Explain the IEEE standard for floating point number representationJune2013

ity
2. A. Draw the hardware implementation of Booths multiplication algorithm.June 2012
B. Trace the steps in the above implementation to multiply -5x-4

.c
C. Illustrate the steps for non restoring division algorithm on the following data:
w
dividend=1011 divisor=0101 ? Dec 2013

w
w
3. A In carry look a head addition explain generating Gi and propagating Pi functions for
stage I with the help of Boolean expression for Gi and P
B.Performance signed multiplication of numbers -12and -11 using both multiplication

algorithm represent the numbers in 5-bit including sign bit .Give booth multiplier
recording table that is used in the above multiplication
C. Perform division of numbers 8 by 3 using non-restoring division algorithms June

2014
4. A Explain booths algorithm Multiply 01110(+14) and 11011(-5) using Booths
om
multiplication July 2 0 13
B.Write the algorithm for binary division using restoring division method
c
C. List the rules for addition, subtraction, multiplication and division of floating point
p.
Numbers June 2012
ou
5. a. Draw and explai n t he single-b u s o r g a nizati o n of tha data pat h in s i de a
proce s s or. Jan 20 1 4
gr
b. W r ite the co n t rol sequen c e for an un -c o n di t i o n al branch instruction.
ts
c . Draw the b l ock diagram of the control u n i t o r g a nization.
en
UNIT-VIII: MULTICORES, MULTIPROCESSORS, AND CLUSTERS

ud
1. A. Write and explain the control sequences for execution of following instruction
st
Add(R3),R1
B. With neat diagram, ex plain 3 bus organization and write control sequence for the
ity
instruction
.c
Add R1,R2,R3.
C. Differentiate between hardwired and micro programmed control June 2013

w
w
2. A. Draw a figure of single bus organization of the processor unit

w
B. list the actions needed to execute the instruction Add R1,(R3). Write the sequence of
control steps to perform actions for a single bus structure. Explain the steps.
C. Compare hardwired control unit with micro programmed control unit. Dec 2013

3.A. Draw the block diagram of the three-bus organization of data path, which provides
multiple internal paths to enable several transfer to take place in parallel. Lable the
registers and fuctional components of the processor and their connection to respective bus
of data path July 2013
om
5. A .Write and explain the control sequences for the execution of an unco nditional
c
branch instruction
p.
B. Explain with block diagram the basic organization of a microprogrammed control
ou
uni t
C. What are modifications required in the basic organization of a microprogrammed
gr
control unit to support conditional branching in the microprogram June 2012
ts
6. List out differences between shared memory multiprocessor and cluster July 2013
en
7. a.Explain in brief about multiprocessor systems. Jan 2014
ud
b. Explain about memory organization in multiprocessors.
c. What are the classifications of parallel processing.

st
ity
.c
w
w
w

VTU QUESTION BANK SOLUTION
UNIT-I: BASIC STRUCTURE OF COMPUTERS

1. With a neat diagram explain different processor registers June 2012(8M)
om
ANS.
Transfers between the memory and the processor are started by sending the address of the
c
memory location to be accessed to the memory unit and issuing the appropriate control signals.
p.
The data are then transferred to or from the memory.
ou
MEMORY
gr
ts
en
ud
MAR
st
MDR CONTROL
ity
R0
.c
PC R1
w
A LU

w
IR
w
Fig b : Connections between the processor and the memory
Dept Of CSE, SJBIT Page 1

The fig shows how memory & the processor can be connected. In addition to the ALU &
the control circuitry, the processor contains a number of registers used for several different
purposes.
The instruction register (IR):- Holds the instructions that is currently being executed. Its
om
output is available for the control circuits which generates the timing signals that control the
various processing elements in one execution of instruction.
The program counter PC:- This is another specialized register that keeps track of execution of
c
a program. It contains the memory address of the next instruction to be fetched and executed.
p.
Besides IR and PC, there are n-general purpose registers R0 through Rn-1.
ou
The other two registers which facilitate communication with memory are: -
1. MAR (Memory Address Register):- It holds the address of the location to be
gr
accessed.
2. MDR (Memory Data Register):- It contains the data to be written into or read out of
ts
the address location.
Operating steps are
en
1. Programs reside in the memory & usually get these through the I/P unit.
2. Execution of the program starts when the PC is set to point at the first instruction of the
ud
program.
3. Contents of PC are transferred to MAR and a Read Control Signal is sent to the memory.
4. After the time required to access the memory elapses, the address word is read out of the
st
memory and loaded into the MDR.

5. Now contents of MDR are transferred to the IR & now the instruction is ready to be
ity
decoded and executed.

6. If the instruction involves an operation by the ALU, it is necessary to obtain the required
operands.
.c
7. An operand in the memory is fetched by sending its address to MAR & Initiating a read
cycle.
w
8. When the operand has been read from the memory to the MDR, it is transferred from
MDR to the ALU.
w
9. After one or two such repeated cycles, the ALU can perform the desired operation.
10. If the result of this operation is to be stored in the memory, the result is sent to MDR.
w
11. Address of location where the result is stored is sent to MAR & a write cycle is initiated.
12. The contents of PC are incremented so that PC points to the next instruction that is to be
executed.

generation of computers Dec 2013, Jan 2012(10M)

Performance: -
The total time required to execute the program is elapsed time is a measure of the
om
performance of the entire computer s ystem. It is affected by the speed of the processor, the disk
and the printer. The time needed to execute a instruction is called the processor time.
Just as the elapsed time for the execution of a program depends on all units in a
c
computer system, the processor time depends on the hardware involved in the execution of
p.
individual machine instructions. This hardware comprises the processor and the memory which
are usually connected by the bus The processor and relatively small cache memory can be
ou
fabricated on a single IC chip. The internal speed of performing the basic steps of instruction
processing on chip is very high and is considerably faster than the speed at which the instruction
and data can be fetched from the main memory. A program will be executed faster if the
gr
movement of instructions and data between the main memory and the processor is minimized,
which is achieved by using the cache.
ts
Processor clock: -
en
Processor circuits are controlled by a timing signal called clock. The clock designer the
regular time intervals called clock cycles. To execute a machine instruction the processor divides
the action to be performed into a sequence of basic steps that each step can be completed in one
ud
clock cycle. The length P of one clock cycle is an important parameter that affects the processor
performance. Processor used in todays personal computer and work station has a clock rates that
range from a few hundred million to over a billion cycles per second.
st
Pipelining and super scalar operation: -

ity
We assume that instructions are executed one after the other. Hence the value of S is the
total number of basic steps, or clock cycles, required to execute one instruction. A substantial
.c
improvement in performance can be achieved by overlapping the execution of successive

instructions using a technique called pipelining.
w
Consider Add R1 R2 R3This adds the contents of R1 & R2 and places the sum into R3. The
w
contents of R1 & R2 are first transferred to the inputs of ALU. After the addition operation is
performed, the sum is transferred to R3. The processor can read the next instruction from the
w
memory, while the addition operation is being performed. Then of that instruction also uses, the
ALU, its operand can be transferred to the ALU inputs at the same time that the add instructions
is being transferred to R3. In the ideal case if all instructions are overlapped to the maximum
degree possible the execution proceeds at the rate of one instruction completed in each clo ck

cycle. Individual instructions still require several clock cycles to complete. But for the purpose of
computing T, effective value of S is 1.
Clock rate:- These are two possibilities for increasing the clock rate R.
1. Improving the IC technology makes logical circuit faster, which reduces the time of
execution of basic steps. This allows the clock period P, to be reduced and the clock rate
om
R to be increased.
2. Reducing the amount of processing done in one basic step also makes it possible to
reduce the clock period P. however if the actions that have to be performed by an
c
instructions remain the same, the number of basic steps needed may increase.
p.
1
n n
SPEC rating = i=1 SPECi

ou
Where n = number of programs in suite. Since actual execution time is measured the SPEC
rating is a measure of the combined effect of all factors affecting performance, including the
gr
compiler, the OS, the processor, the memory of comp being tested.
ts
3. Discuss two ways in which b yte addresses are assigned June 2012(10M)
en
Ans. BIG-ENDIAN AND LITTLE-ENDIAN ASIGNMENTS:-
There are two ways that byte addresses can be assigned across words, as shown in fig b.
ud
The name big-endian is used when lower b yte addresses are used for the more significant b ytes
(the leftmost bytes) of the word. The name little-endian is used for the opposite ordering, where
the lower byte addresses are used for the less significant bytes (the rightmost bytes) of the word.
st
In addition to specif ying the address ordering of bytes within a word, it is also necessary
ity
to specify the labeling of bits within a byte or a word. The same ordering is also used for labeling
bits within a byte, that is, b7, b6, ., b0, from left to right.
.c
WORD ALIGNMENT:-
w
In the case of a 32-bit word length, natural word boundaries occur at addresses 0, 4, 8,
, as shown in above fig. We say that the word locations have aligned addresses . in general,
w
words are said to be aligned in memory if they begin at a byte address that is a multiple of the
number of bytes in a word. The memory of b ytes in a word is a power of 2. Hence, if the word
w
length is 16 (2 b ytes), aligned words begin at byte addresses 0,2,4,, and for a word length of
64 (23 bytes), aligned words begin at bytes addresses 0,8,16 .

4. Explain the different functional units of a computer with a neat block diagram
Dec 2013(10M)
Ans. Functional unit: -
A computer consists of five functionally independent main parts input, memory,

arithmetic logic unit (ALU), output and control unit.
om
Input A LU
c
p.
Memory
ou
IO/O
utput Processor
Control Unit
gr
Fig a : Functional units of computer
ts
Input device accepts the coded information as source program i.e. high level language.
en
This is either stored in the memory or immediately used by the processor to perform the desired
operations. The program stored in the memory determines the processing steps. Basically the
computer converts one source program to an object program. i.e. into machine language. Finally
ud
the results are sent to the outside world through output device. All of these actions are
coordinated by the control unit.
st
Input unit: -
ity
The source program/high level language program/coded information/simply data is fed

to a computer through input devices keyboard is a most common type. Whenever a key is
pressed, one corresponding word or number is translated into its equivalent binary code over a
.c
cable & fed either to memory or processor. Joysticks, trackballs, mouse, scanners etc are other
input devices.
w
Memory unit: -
w
Its function into store programs and data. It is basically to two types
w
Primary memory
Secondary memory

1. Primary memory: - Is the one exclusively associated with the processor and operates at the
electronics speeds programs must be stored in this memory while they are being executed. The
memory contains a large number of semiconductors storage cells. Each capable of storing one bit
of information. These are processed in a group of fixed site called word. To provide easy
access to a word in memory, a distinct address is associated with each word location. Addresses
are numbers that identify memory location. Number of bits in each word is called word length of
om
the computer. Programs must reside in the memory during execution. Instructions and data can
be written into the memory or read out under the control of processor. Memory in which any
location can be reached in a short and fixed amount of time after specif ying its address is called
random-access memory (RAM). The time required to access one word in called memory
c
access time. Memory which is only readable by the user and contents of which cant be altered is
p.
called read only memory (ROM) it contains operating system.
ou
2 Secondary memory: - Is used where large amounts of data & programs have to be stored,
particularly information that is accessed infrequently. Examples: - Magnetic disks & tapes,
optical disks (ie CD-ROMs), floppies etc.,
gr
Arithmetic logic unit (ALU):-
ts
Most of the computer operators are executed in ALU of the processor like addition,
subtraction, division, multiplication, etc. the operands are brought into the ALU from memory
en
and stored in high speed storage elements called register. Then according to the instructions the
operation is performed in the required sequence. The control and the ALU are may times
faster than other devices connected to a computer system. This enables a single processor to
ud
control a number of external devices such as key boards, displays, magnetic and optical disks,
sensors and other mechanical controllers.
st
Output unit:-
These actually are the counterparts of input unit. Its basic function is to send the
ity
processed results to the outside world.Examples:- Printer, speakers, monitor etc.
Control unit:-
.c
It effectively is the nerve center that sends signals to other units and senses their states.
w
The actual timing signals that govern the transfer of data between input unit, processor, memory
and output unit are generated by the control unit.
w
w

5. Write the basic performance equation. explain the role of each of the parameters in
the equation on the performance of the computer. June 2013(12M)
Ans. Basic performance equation: -
We now focus our attention on the processor time component of the total elapsed time.
Let T be the processor time required to execute a program that has been prepared in some high-
om
level language. The compiler generates a machine language object program that corresponds to
the source program. Assume that complete execution of the program requires the execution of N
machine cycle language instructions. The number N is the actual number of instruction execution
c
and is not necessarily equal to the number of machine cycle instructions in the object program.
Some instruction may be executed more than once, which in the case for instructions inside a
p.
program loop others may not be executed all, depending on the input data used.
ou
Suppose that the average number of basic steps needed to execute one machine cycle
instruction is S, where each basic step is completed in one clock cycle. If clock rate is R cycles
per second, the program execution time is given by
gr
NS
T = this is often referred to as the basic performance equation.
R
ts
We must emphasize that N, S & R are not independent parameters changing one may
en
affect another. Introducing a new feature in the design of a processor will lead to improved
performance only if the overall result is to reduce the value of T.
ud
6. With a neat diagram ,discuss the basic operational concepts of a computer

July 2013(10M)
st
Basic Operational Concepts

ity
Instructions take a vital role for the proper working of the computer.
An appropriate program consisting of a list of instructions is stored in the memory so that
.c
the tasks can be started.

The memory brings the Individual instructions into the processor, which executes the
w
specified operations.
w
Data which is to be used as operands are moreover also stored in the memory.
w

Example:
Add LOCA, R0
This instruction adds the operand at memory location LOCA to the operand
which will be present in the Register R0.
om
The above mentioned example can be written as follows:
7. List the different systems used to repres ent a signed number Jan 2014(10M)
c
Ans: 1)Sign and magnitude
p.
2) 1s complement
ou
3) 2s complement
8. Explain with necessary block diagram the basic functional unit of a computer.
gr
Jan 2014,Jan 2012( 1 0 M ) ts
Ans. First Generation (1940-1956) Vacuum Tubes
The first computers used vacuum tubes for circuitry and magnetic drums for memory, and were
en
often enormous, taking up entire rooms. They were very expensive to operate and in addition to
using a great deal of electricity, generated a lot of heat, which was often the cause of
ud
malfunctions.
First generation computers relied on machine language, the lowest-level programming language
st
understood by computers, to perform operations, and they could only solve one problem at a
time. Input was based on punched cards and paper tape, and output was displayed on printouts.
ity
.c
Second Generation (1956-1963) Transistors

Transistors replaced vacuum tubes and ushered in the second generation of computers. The
w
transistor was invented in 1947 but did not see widespread use in computers until the late 1950s.
w
The transistor was far superior to the vacuum tube, allowing computers to become smaller,
w
faster, cheaper, more energy-efficient and more reliable than their first-generation predecessors.
Though the transistor still generated a great deal of heat that subjected the computer to damage,

it was a vast improvement over the vacuum tube. Second-generation computers still relied on
punched cards for input and printouts for output.
Third Generation (1964-1971) Integrated Circuits
om
The development of the integrated circuit was the hallmark of the third generation of computers.
Transistors were miniaturized and placed on silicon chips, called semiconductors, which
drastically increased the speed and efficiency of computers.
c
Instead of punched cards and printouts, users interacted with third generation computers
p.
through keyboards and monitorsand interfaced with an operating system, which allowed the
device to run many different applications at one time with a central program that monitored the
ou
memory. Computers for the first time became accessible to a mass audience because they were
smaller and cheaper than their predecessors.
gr
ts
Fourth Generation (1971-Present) Microprocessors
The microprocessor brought the fourth generation of computers, as thousands of integrated
en
circuits were built onto a single silicon chip. What in the first generation filled an entire room
could now fit in the palm of the hand. The Intel 4004 chip, developed in 1971, located all the
ud
components of the computerfrom the central processing unit and memory to input/output
controlson a single chip.
st
ity
Fifth Generation (Present and Beyond) Artificial Intelligence

Fifth generation computing devices, based on artificial intelligence, are still in development,
.c
though there are some applications, such as voice recognition, that are being used today. The use
w
of parallel processing and superconductors is helping to make artificial intelligence a

reality. Quantum computation and molecular and nanotechnology will radically change the face
w
of computers in years to come. The goal of fifth-generation computing is to develop devices that
w
respond to natural language input and are capable of learning and self-organization.

UNIT-II- MACHINE INSTRUCTIONS AND PROGRAMS CONTD.:
1. What is an addressing mode .Explain different generic addressing modes with an

example for each June 2012/July 2013,Jan 2014.
om
Ans.Addressing modes: In general, a program operates on data that reside in the computers
memory. These data can be organized in a variety of ways. If we want to keep track of students
names, we can write them in a list. Pro grammers use organizations called data structures to
c
represent the data used in computations. These include lists, linked lists, arrays, queues, and so
on .
p.
Programs are normally written in a high-level language, which enables the programmer
ou
to use constants, local and global variables, pointers, and arra ys. The different ways in which the
location of an operand is specified in an instruction are referred to as addressing modes.
gr
Name Assembler syntax
ts Addressing function
en
Immediate # Value Operand = Value
Register Ri EA = Ri
ud
Absolute (Direct) LOC EA = LOC
Indirect (R i ) EA = [Ri]
st
(LOC) EA = [LOC]
ity
Index X(Ri) EA = [Ri] + X

.c
Base with index (Ri, Rj) EA = [Ri] + [Rj]
Base with index X (Ri, Rj) EA = [Ri] + [Rj] + X

w
and offset
w
Relative X(PC) EA = [PC] + X

w
Autoincrement (Ri)+ EA = [Ri]; Increment Ri
Autodecrement -(Ri) Decrement Ri; EA = [Ri]
EA = effective address

Value = a signed number
2.Define subroutine. Explain subroutine linkage using a link register (8M) Dec 2013
Ans. Subroutines
In a given program, it is often necessary to perform a particular subtask many times on
om
different data-values. Such a subtask is usually called a subroutine. For example, a subroutine
may evaluate the sine function or sort a list of values into increasing or decreasing order. It is
possible to include the block of instructions that constitute a subroutine at every place where it is
needed in the program. However, to save space, only one copy of the instructions that constitute
c
the subroutine is placed in the memory, and any program that requires the use of the subroutine
p.
simply branches to its starting location. When a program branches to a subroutine we say that it
is calling the subroutine. The instruction that performs this branch operation is named a Call
ou
instruction.
After a subroutine has been executed, the calling program must resume execution,
gr
continuing immediately after the instruction that called the subroutine. The subroutine is said to
return to the program that called it by executing a Return instruction.
ts
The way in which a computer makes it possible to call and return from subroutines is
referred to as its subroutine linkage method. The simplest subroutine linkage method is to save
en
the return address in a specific location, which may be a register dedicated to this function. Such
a register is called the link register. When the subroutine completes its task, the Return
instruction returns to the calling program by branching indirectly through the link register.
ud
The Call instruction is just a special branch instruction that performs the following operations
st

ity

.c
w
Memory Memory
w

w

20 0 Call SUB 10 00 first instruction
20 4 next instruction . Return
100 0
om
204
c
PC
p.
ou
Link
204
gr
Call Return
ts
Subroutine linkage using a link registers
3.Explain the shift and rotate operation with examples(8M)

en
June 2013
Ans. SHIFT AND ROTATE INSTRUCTIONS

ud
There are many applications that require the bits of an operand to be shifted right or left
some specified number of bit positions. The details of how the shifts are performed depend on
whether the operand is a signed number or some more general binary-coded information. For
st
general operands, we use a logical shift. For a number, we use an arithmetic shift, which
preserves the sign of the number.
ity
Logical shifts:-
.c
Two logical shift instructions are needed, one for shifting left (LShiftL) and another for
shifting right (LShiftR). These instructions shift an operand over a number of bit positions
w
specified in a count operand contained in the instruction. The general form of a logical left shift
instruction is
w
LShiftL count, dst

w
(a) Logical shift left LShiftL #2, R0
R0

0 0 1 1 1 0 . . . 0 1 1
om
C
before :
c
p.
1 1 1 0 . . . 0 1 1 00
ou
after:
gr
(b) Logical shift right
tsLShiftR #2, R0
en
R0 C
ud
0 1 1 1 0 . . . 0 1 1 0
Before:
st
1
ity
After: 0 0 0 1 1 1 0 . . . 0
.c
( c) Arithmetic shift right AShiftR #2, R0

w
w
w
R0 C

Before:
1 0 0 1 1 . . . 0 1 0 0
After: 1 1 1 0 0 1 1 . . . 0 1
Rotate Operations:-
om
In the shift operations, the bits shifted out of the operand are lost, except for the last bit
shifted out which is retained in the Carry flag C. To preserve all bits, a set of rotate instructions
can be used. They move the bits that are shifted out of one end of the operand back into the other
c
end. Two versions of both the left and right rotate instructions are usually provided. In one
p.
version, the bits of the operand are simply rotated. In the other version, the rotation includes the
C flag.
ou
(a) Rotate left without carry RotateL #2, R0
(b) Rotate left with carry RotateLC #2, R0
gr
4. What is the need for an addressing mode? Explain the following addressing modes with
examples: immediate, direct, indirect, index, relative(8M) Dec 2013/July 2013
ts
en
Ans. Immediate mode The operand is given explicitly in the instruction.
For example, the instruction

ud
Move 200immediate, R0 Places the value 200 in register R0. Clearly, the
Immediate mode is only used to specify the value of a source operand. Using a subs cript to
st
denote the Immediate mode is not appropriate in assembly languages. A common convention is
to use the sharp sign (#) in front of the value to indicate that this value is to be used as an
ity
immediate operand. Hence, we write the instruction above in the form
Move #200, R0
.c
Direct mode:
w
Register mode - The operand is the contents of a processor register; the name (address) of the
w
register is given in the instruction.
Absolute mode The operand is in a memory location; the address of this location is given
w
explicitly in the instruction. (In some assembly languages, this mode is called Direct).
The instruction
Move LOC, R2

Processor registers are used as temporary storage locations where the data is a register
are accessed using the Register mode. The Absolute mode can represent global variables in a
program. A declaration such as
Integer A, B;
5. a)What is subroutine linkage?how are parameters passed to subroutines(8M) June 2014
om
In a given program, it is often necessary to perform a particular subtask many times on
different data-values. Such a subtask is usually called a subroutine. For example, a subroutine
may evaluate the sine function or sort a list of values into increasing or decreasing order. It is
c
possible to include the block of instructions that constitute a subroutine at every place where it is
p.
needed in the program. However, to save space, only one copy of the instructions that constitute
the subroutine is placed in the memory, and any program that requires the use of the subroutine
ou
simply branches to its starting location. When a program branches to a subroutine we say that it
is calling the subroutine. The instruction that performs this branch operation is named a C all
instruction.
gr
After a subroutine has been executed, the calling program must resume execution,
continuing immediately after the instruction that called the subroutine. The subroutine is said to
ts
return to the program that called it by executing a Return instruction.
en
The way in which a computer makes it possible to call and return from subroutines is
referred to as its subroutine linkage method. The simplest subroutine linkage method is to save
the return address in a specific location, which may be a register dedicated to this function. Such
ud
a register is called the link register. When the subroutine completes its task, the Return
instruction returns to the calling program by branching indirectly through the link register.
st
The Call instruction is just a special branch instruction that performs the following operations
ity

.c

w
w
Memory Memory
w

20 0 Call SUB 10 00 first instruction
20 4 next instruction . Return
om
100 0
c
204
p.
PC
ou
Link
gr
204
Call
ts Return
en
Subroutine linkage using a link registers
b.What is stack frame? Explain. (6M)

ud
THE STACK FRAME:-
Now, observe how space is used in the stack in the example. During execution of the
st
subroutine, six locations at the top of the stack contain entries that are needed by the subroutine.
These locations constitute a private workspace for the subroutine, created at the time the
ity
subroutine is entered and freed up when the subroutine returns control to the calling program.
Such space is called a stack frame.
.c
Fig a A subroutine stack frame example.

Saved [R1]
SP
w
Saved [R0] (stack pointer)

w
Stack
Localvar3
w
Localvar2 frame
for
Localvar1
FP called
Saved [FP]

(frame pointer) subroutine
Old TOS
fig b shows an example of a commonly used layout for information in a stack frame. In
om
addition to the stack pointer SP, it is useful to have another pointer register, called the frame
pointer (FP), for convenient access to the parameters passed to the subroutine and to the local
memory variables used by the subroutine. These local variables are only used within the
subroutine, so it is appropriate to allocate space for them in the stack frame associated with the
c
subroutine. We assume that four parameters are passed to the subroutine, three local variables are
p.
used within the subroutine, and registers R0 and R1 need to be saved because they will also be
used within the subroutine.
ou
The pointers SP and FP are manipulated as the stack frame is built, used, and dismantled
for a particular of the subroutine. We begin by assuming that SP point to the old top-of-stack
(TOS) element in fig b. Before the subroutine is called, the calling p rogram pushes the four
gr
parameters onto the stack. The call instruction is then executed, resulting in the return address
being pushed onto the stack. Now, SP points to this return address, and the first instruction of the
ts
subroutine is about to be executed. This is the point at which the frame pointer FP is set to
contain the proper memory address. Since FP is usually a general-purpose register, it may
en
contain information of use to the Calling program. Therefore, its contents are saved by pushing
them onto the stack. Since the SP now points to this position, its contents are copied into FP.
ud
Thus, the first two instructions executed in the subroutine are
Move FP, -(SP)

st
Move SP, FP
ity
After these instructions are executed, both SP and FP point to the saved FP contents.
Subtract #12, SP
.c
Finally, the contents of processor registers R0 and R1 are saved by pushing them onto
w
the stack. At this point, the stack frame has been set up.
w
The subroutine now executes its task. When the task is completed, the subroutine pops
the saved values of R1 and R0 back into those registers, removes the local variables from the
w
stack frame by executing the instruction.
A dd #12, SP

And pops the saved old value of FP back into FP. At this point, SP points to the return
address, so the Return instruction can be executed, transferring control back to the calling
program.
6. Discuss briefly encoding of machine instructions (8M) July 2013
om
Computer instructions are the basic components of a machine language program. They are also
known as macrooperations, since each one is comprised of a sequences of microoperations.
Each instruction initiates a sequence of microoperations that fetch operands from registers or
memory, possibly perform arithmetic, logic, or shift operations, and store results in registers or
c
memory.
p.
Instructions are encoded as binary instruction codes. Each instruction code contains of
ou
a operation code, or opcode, which designates the overall purpose of the instruction (e.g. add,
subtract, move, input, etc.). The number of bits allocated for the opcode determined how many
different instructions the architecture supports.
gr
In addition to the opcode, many instructions also contain one or more operands, which indicate
where in registers or memory the data required for the operation is located. For example,
ts
and add instruction requires two operands, and a not instruction requires one.
en
15 12 11 6 5 0
ud
+-----------------------------------+
| Opcode | Operand | Operand |
+-----------------------------------+
st
The opcode and operands are most often encoded as unsigned binary numbers in order to
ity
minimize the number of bits used to store them. For example, a 4-bit opcode encoded as a binary
number could represent up to 16 different operations.
.c
The control unit is responsible for decoding the opcode and operand bits in the instruction
register, and then generating the control signals necessary to drive all other hardware in the CPU
w
to perform the sequence of microoperations that comprise the instruction.

w
7. Big-Endian and Little-Endian assignment,explain with necessary figure. Jan 2014

w
Big Endian
In big endian, you store the most significant byte in the smallest address. Here's how it would
look:

Address Value
100 0 90
100 1 AB
om
100 2 12
c
100 3 CD
p.
ou
Little Endian
In little endian, you store the least significant byte in the smallest address. Here's how it would
gr
look:
ts Address Value
en
100 0 CD
ud
100 1 12
100 2 AB
st
ity
100 3 90
Notice that this is in the reverse order compared to big endian. To remember which is which,
.c
recall whether the least significant byte is stored first (thus, little endian) or the most significant
w
byte is stored first (thus, big endian).

w
Notice I used "byte" instead of "bit" in least significant bit. I sometimes abbreciated this as LSB
w
and MSB, with the 'B' capitalized to refer to byte and use the lowercase 'b' to represent bit. I only
refer to most and least significant byte when it comes to endianness.

8. Explain logical and arithmetic shift instructions with an example. Jan 2014
Ans: The restriction that an instruction must occupy only one word has
led to a style of computers that have become known as reduced instruction
set computer (R ISC). The RISC approach introduced other restrictions, such
om
as that all manipulation of data must be done on operands that are already in
processor registers. This restriction means that the above addition would need
a two-instruction sequ ence
c
Move (R3), R1
p.
Add R 1, R 2
ou
gr
If the Add instruction only has to specify the two registers, it will
need just a portion of a 32-bit word. So, we may provide a more powerful
instruction that uses three operands
ts
Add R1, R2, R3
en
Which performs the operation

ud
R3 [R1] + [R2]
st
ity
A possible format for such an instruction in shown in fig c. Of

course, the processor has to be able to deal with such three-operainstructions.
.c
In an instruction set where all arithmetic and logical operations use only
register operands, the only memory references are made to load/store the
w
operands into/from the processor registers.

w
RISC-t ype instruction sets t ypically have fewer and less complex
w
instructions than CISC-type sets. We will discuss the relative merits of R ISC
and C ISC approaches in

UNIT-III- INPUT/OUTPUT ORGANIZATION

1..Define and explain briefly the following
i) interrupt ii)vectored interrupt ii)interrupt nesting iv)an exception and give two
examples (6M) July 2013
om
Ans. Interrupt
We pointed out that an I/O device requests an interrupt by activating a bus line called
interrupt-request. Most computers are likely to have several I/O devices that can request an
c
interrupt. A single interrupt-request line may be used to serve n devices as depicted. All devices
p.
are connected to the line via switches to ground. To request an interrupt, a device closes its
associated switch. Thus, if all interrupt-request signals INTR1 to INTRn are inactive, that is, if all
ou
switches are open, the voltage on the interrupt-request line will be equal to Vdd. This is the
inactive state of the line. Since the closing of one or more switches will cause the line voltage to
drop to 0, the value of INTR is the logical OR of the requests from individual devices, that is,
gr
INTR = INTR1 + +INTRn
ts
It is customary to use the complemented form, INTR , to name the interrupt-request signal on
the common line, because this signal is active when in the low-voltage state.
en
Vectored Interrupts:-
To reduce the time involved in the polling process, a device requesting an interrupt may
ud
identify itself directly to the processor. Then, the processor can immediately start executing the
corresponding interrupt-service routine. The term vectored interrupts refers to all interrupt-
st
handling schemes based on this approach.
A device requesting an interrupt can identify itself by sending a special code to the
ity
processor over the bus. This enables the processor to identify individual devices even if they
share a single interrupt-request line. The code supplied by the device may represent the starting
.c
address of the interrupt-service routine for that device. The code length is typically in the range
of 4 to 8 bits. The remainder of the address is supplied by the processor based on the area in its
w
memory where the addresses for interrupt-service routines are located.
This arrangement implies that the interrupt-service routine for a given device must
w
always start at the same location. The programmer can gain some flexibility by storing in this
w
location an instruction that causes a branch to the appropriate routine.

Interrupt Nesting: -
Interrupts should be disabled during the execution of an interrupt-service routine, to

ensure that a request from one device will not cause more than one interruption. The same
arrangement is often used when several devices are involved, in which case execution of a given
interrupt-service routine, once started, always continues to completion before the processor
om
accepts an interrupt request from a second device. Interrupt-service routines are typically short,
and the delay they may cause is acceptable for most simple devices.
For some devices, however, a long delay in responding to an interrupt request may lead
c
to erroneous operation. Consider, for example, a computer that keeps track of the time of day
using a real-time clock. This is a device that sends interrupt requests to the processor at regular
p.
intervals. For each of these requests, the processor executes a short interrupt-service routine to
increment a set of counters in the memory that keep track of time in seconds, minutes, and so on.
ou
Proper operation requires that the delay in responding to an interrupt request from the real-time
clock be small in comparison with the interval between two successive requests. To ensure that
this requirement is satisfied in the presence of other interrupting devices, it may be necessary to
gr
accept an interrupt request from the clock during the execution of an interrupt-service routine for
another device.
ts
This example suggests that I/O devices should be organized in a priority structure. An
en
interrupt request from a high-priority device should be accepted while the processor is servicing
another request from a lower-priority device.
ud
A multiple-level priority organization means that during execution of an interrupt-

service routine, interrupt requests will be accepted from some devices but not from others,
depending upon the devices priority. To implement this scheme, we can assign a priority level
st
to the processor that can be changed under program control. The priority level of the processor is
the priority of the program that is currently being executed. The processor accepts interrupts only
ity
from devices that have priorities higher than its own.
The processors priority is usually encoded in a few bits of the processor status word. It
.c
can be changed by program instructions that write into the PS. These are privileged instructions,
which can be executed only while the processor is running in the supervisor mode. The processor
w
is in the supervisor mode only when executing operating s ystem routines. It switches to the user
mode before beginning to execute application programs. Thus, a user program cannot
w
accidentally, or intentionally, change the priority of the processor and disrupt the systems
operation. An attempt to execute a privileged instruction while in the user mode leads to a
w
special type of interrupt called a privileged instruction.
A multiple-priority scheme can be implemented easily by using separate interrupt-

request and interrupt-acknowledge lines for each device, as shown in figure. Each of the
interrupt-request lines is assigned a different priority level. Interrupt requests received over these

lines are sent to a priority arbitration circuit in the processor. A request is accepted only if it has a
higher priority level than that currently assigned to the processor.
2.Explain in with the help of a diagram the working of daisy chain with multiple priority
levels and multiple devices in each level (8M) Dec 2013,JAN 2014
Simultaneous Requests:
om
Let us now consider the problem of simultaneous arrivals of interrupt requests from two
or more devices. The processor must have some means of deciding which requests to serv ice
first. Using a priority scheme such as that of figure, the solution is straightforward. The processor
c
simply accepts the requests having the highest priority.
p.
Polling the status registers of the I/O devices is the simplest such mechanism. In this
ou
case, priority is determined by the order in which the devices are polled. When vectored
interrupts are used, we must ensure that only one device is selected to send its interrupt vector
code. A widely used scheme is to connect the devices to form a daisy ch ain, as shown in figure
gr
3a. The interrupt-request line INTR is common to all devices. The interrupt-acknowledge line,
INTA, is connected in a daisy-chain fashion, such that the INTA signal propagates serially
ts
through the devices.
en
INTR
ud
st
Device 1 Device 2 Device n

ity
INTA (3.a) Daisy chain
When several devices raise an interrupt request and the INTR line is activated, the
.c
processor responds by setting the INTA line to 1. This signal is received by device 1. Device 1
passes the signal on to device 2 only if it does not require any service. If device 1 has a pending
w
request for interrupt, it blocks the INTA signal and proceeds to put its identifying code on the
data lines. Therefore, in the daisy-chain arrangement, the device that is electrically closest to the
w
processor has the highest priority. The second device along the chain has second highest prio rity,
and so on.
w
The scheme in figure 3.a requires considerably fewer wires than the individual
connections in figure 2. The main advantage of the scheme in figure 2 is that it allows the
processor to accept interrupt requests from some devices but not from others, depending upon
their priorities. The two schemes may be combined to produce the more general structure in

figure 3b. Devices are organized in groups, and each group is connected at a different priority
level. Within a group, devices are connected in a daisy chain. This organization is used in many
computer systems.
3. Discuss the different schemes available to disable and enable the interrupts(8M)
om
June 2013 /July 2013
Ans. ENABLING AND DISABLING INTERRUPTS:
c
The facilities provided in a computer must give the programmer complete control over
p.
the events that take place during program execution. The arrival of an interrupt request from an
external device causes the processor to suspend the execution of one program and start the
ou
execution of another. Because interrupts can arrive at any time, they may alter the sequence of
events from the envisaged by the programmer. Hence, the interruption of program execution
must be carefully controlled.
gr
Let us consider in detail the specific case of a single interrupt request from one device.
When a device activates the interrupt-request signal, it keeps this signal activated until it learns
ts
that the processor has accepted its request. This means that the interrupt-request signal will be
active during execution of the interrupt-service routine, perhaps until an instruction is reached
en
that accesses the device in question.
The first possibility is to have the processor hardware ignore the interrupt-request line
ud
until the execution of the first instruction of the interrupt-service routine has been completed.
Then, by using an Interrupt-disable instruction as the first instruction in the interrupt-service
routine, the programmer can ensure that no further interruptions will occur until an Interrupt-
st
enable instruction is executed. Typically, the Interrupt-enable instruction will be the last
instruction in the interrupt-service routine before the Return-from-interrupt instruction. The
ity
processor must guarantee that execution of the Return-from-interrupt instruction is completed

before further interruption can occur.
.c
The second option, which is suitable for a simple processor with only one interrupt-
request line, is to have the processor automatically disable interrupts before starting the
w
execution of the interrupt-service routine. After saving the contents of the PC and the processor
status register (PS) on the stack, the processor performs the equivalent of executing an Interrupt-
w
disable instruction. It is often the case that one bit in the PS register, called Interrupt-enable,
indicates whether interrupts are enabled.
w
interrupt-handling circuit responds only to the leading edge of the signal. Such a line is said to be
edge-triggered.Before proceeding to study more complex aspects of interrupts, let us summarize

the sequence of events involved in handling an interrupt request from a single device. Assuming
that interrupts are enabled, the following is a typical scenario.

3. Interrupts are disabled by changing the control bits in the PS (except in the case of edge-
om
triggered interrupts).
c
p.
4. Discuss the different schemes available to disable and enable the interrupts(8M)Dec 2014
ou
Ans. ENABLING AND DISABLING INTERRUPTS:
The facilities provided in a computer must give the programmer complete control over
gr
the events that take place during program execution. The arrival of an interrupt request from an
external device causes the processor to suspend the execution of one program and start the
ts
execution of another. Because interrupts can arrive at any time, they may alter t he sequence of
events from the envisaged by the programmer. Hence, the interruption of program execution
en
must be carefully controlled.
Let us consider in detail the specific case of a single interrupt request from one device.
ud
When a device activates the interrupt-request signal, it keeps this signal activated until it learns
that the processor has accepted its request. This means that the interrupt-request signal will be
active during execution of the interrupt-service routine, perhaps until an instruction is reached
st
that accesses the device in question.
The first possibility is to have the processor hardware ignore the interrupt-request line
ity
until the execution of the first instruction of the interrupt-service routine has been completed.
Then, by using an Interrupt-disable instruction as the first instruction in the interrupt-service
.c
routine, the programmer can ensure that no further interruptions will occur until an Interrupt-
enable instruction is executed. Typically, the Interrupt-enable instruction will be the last
w
instruction in the interrupt-service routine before the Return-from-interrupt instruction. The

processor must guarantee that execution of the Return-from-interrupt instruction is completed
w
before further interruption can occur.

w
The second option, which is suitable for a simple processor with only one interrupt-
request line, is to have the processor automatically disable interrupts before starting the
execution of the interrupt-service routine. After saving the contents of the PC and the pro cessor
status register (PS) on the stack, the processor performs the equivalent of executing an Interrupt-

disable instruction. It is often the case that one bit in the PS register, called Interrupt-enable,
indicates whether interrupts are enabled.
interrupt-handling circuit responds only to the leading edge of the signal. Such a line is said to be
edge-triggered.Before proceeding to study more complex aspects of interrup ts, let us summarize
om
the sequence of events involved in handling an interrupt request from a single device. Assuming
that interrupts are enabled, the following is a typical scenario.
c
3. Interrupts are disabled by changing the control bits in the PS (except in the case of edge-
p.
triggered interrupts).
ou
gr
ts
5. Write a short note on any 1 bus arbitration scheme.(5M)June 2013/July 2014
en
Centralized Arbitration:- The bus arbiter may be the processor or a separate unit connected to
the bus. A basic arrangement in which the processor contains the bus arbitration circuitry. In this
case, the processor is normally the bus master unless it grants bus mastership to one of the DMA
ud
controllers. A DMA controller indicates that it needs to become the bus master by activating the
Bus-Request line,The signal on the Bus-Request line is the logical OR of the bus requests from
all the devices connected to it. When Bus-Request is activated, the processor activates the Bus-
st
Grant signal, BG1, indicating to the DMA controllers that they may use the bus when it becomes
free. This signal is connected to all DMA controllers using a daisy-chain arrangement. Thus, if
ity
DMA controller 1 is requesting the bus, it blocks the propagation of the grant signal to other
devices. Otherwise, it passes the grant downstream by asserting BG2. The current bus master
.c
indicates to all device that it is using the bus by activating another open-controller line called
Bus-Busy, BBSY . Hence, after receiving the Bus-Grant signal, a DMA controller waits for Bus-
w
Busy to become inactive, then assumes mastership of the bus. At this time, it activates Bus-Busy
to prevent other devices from using the bus at the same time.
w
7. D r a w the arrangement of a single b u s s t ruct u r e and b r i e f about memory

w
ma p p ed I/O. Jan 2014
A n s : a central control unit and arithmetic logic unit (ALU, which he called the central arithmetic
part) were combined with computer memory and input and output functions to form a stored

program computer.[1] The Report presented a general organization and theoretical model of the
computer, however, not the implementation of that model.[2] Soon designs integrated the control
unit and ALU into what became known as the central processing unit (CPU).
Computers in the 1950s and 1960s were generally constructed in an ad-hoc fashion. For
om
example, the CPU, memory, and input/output units were each one or more cabinets connected by
cables. Engineers used the common techniques of standardized bundles of wires and extended
the concept as backplanes were used to hold printed circuit boards in these early machines.
c
p.
ou
gr
ts
en
ud
7 b) Explain i)inte r r u p t enabling ii)ed g e triggering with resp e c t i n t err u p ts.Jan 2014
A n s: IF (Interrupt Flag) is a system flag bit in the x86 architecture's FLAGS register, which
st
determines whether or not the CPU will handle maskable hardware interrupts.
The bit, which is bit 9 of the FLAGS register, may be set or cleared by programs with sufficient
ity
privileges, as usually determined by the Operating System. If the flag is set to 1, maskable
.c
hardware interrupts will be handled. If cleared (set to 0), such interrupts will be ignored. IF does
not affect the handling of non-maskable interrupts or software interrupts generated by
w
the INTinstruction.
w
w

UNIT-IV- INPUT/OUTPUT ORGANIZATION CONTD.

1. Explain with a block diagram a general 8 bit parallel interface.(8M) June 2013/July 2013
Ans. My-address
RS 2
om
RS 1
Register Status
RS 0 C1
c
select and
Ready
p.
R/W
ou
C2
Accept 16 A general 8-bit parallel interface
gr
The circuit in figure 16 has separate input and output data lines for connection to an I/O device.
A more flexible parallel port is created if the data lines to I/O devices are bidirectional. Figure 17
ts
shows a general-purpose parallel interface circuit that can be configured in a variety of wa ys.
Data lines P7 through P0 can be used for either input or output purposes. For increased
en
flexibility, the circuit makes it possible for some lines to serve as inputs and some lines to serve
as outputs, under program control. The DATAOUT register is connected to these lines via three-
ud
state drivers that are controlled by a data direction register, DDR. The processor can write any 8-
bit pattern into DDR. For a given bit, if the DDR value is 1, the corresponding data line acts as
an output line; otherwise, it acts as an input line.
st
2.With the help of a data transfer signals explain how a real operation is performed using
PCI bus. (8M) Dec 2013
ity
Ans. Peripheral Component Interconnect (PCI) Bus:-

.c
The PCI bus is a good example of a s ystem bus that grew out of the need for
standardization. It supports the functions found on a processor bus bit in a standardized format
w
that is independent of any particular processor. Devices connected to the PCI bus appear to the
processor as if they were connected directly to the processor bus. They are assigned addresses in
w
the memory address space of the processor.

w
The PCI follows a sequence of bus standards that were used primarily in IBM PCs. Early
PCs used the 8-bit XT bus, whose signals closely mimicked those of Intels 80x86 processors.
Later, the 16-bit bus used on the PC At computers became known as the ISA bus. Its extended

32-bit version is known as the EISA bus. Other buses developed in the eighties with similar
capabilities are the Microchannel used in IBM PCs and the NuBus used in Macintosh computers.
The PCI was developed as a low-cost bus that is truly processor independent. Its design
anticipated a rapidly growing demand for bus bandwidth to support high-speed disks and graphic
and video devices, as well as the specialized needs of multiprocessor s ystems. As a result, the
om
PCI is still popular as an industry standard almost a decade after it was first introduced in 1992.
An important feature that the PCI pioneered is a plug-and-play capability for connecting I/O
devices. To connect a new device, the user simply connects the device interface board to the bus.
The software takes care of the rest.
c
p.
Data Transfer:-
ou
In todays computers, most memory transfers involve a burst of data rather than just one
word. The reason is that modern processors include a cache memor y. Data are transferred
gr
between the cache and the main memory in burst of several words each. The words involved in
such a transfer are stored at successive memory locations. When the processor (actually the
cache controller) specifies an address and requests a read operation from the main memory, the
ts
memory responds by sending a sequence of data words starting at that address. Similarly, during
a write operation, the processor sends a memory address followed by a sequence of data words,
en
to be written in successive memory locations starting at the address. The PCI is designed
primarily to support this mode of operation. A read or write operation involving a single word is
ud
simply treated as a burst of length one.
The bus supports three independent address spaces: memory, I/O, and configuration. The
first two are self-explanatory. The I/O address space is intended for use with processors, such as
st
Pentium, that have a separate I/O address space. However, as noted , the system designer may
choose to use memory-mapped I/O even when a separate I/O address space is available. In fact,
ity
this is the approach recommended by the PCI its plug-and-play capability. A 4-bit command that
accompanies the address identifies which of the three spaces is being used in a given data
.c
transfer operation.
The signaling convention on the PCI bus is similar to the one used, we assumed that the
w
master maintains the address information on the bus until data transfer is completed. But, this is
w
not necessary. The address is needed only long enough for the slave to be selected. The slave can
store the address in its internal buffer. Thus, the address is needed on the bus for one clock cycle
w
only, freeing the address lines to be used for sending data in subsequent clock cycles. The result
is a significant cost reduction because the number of wires on a bus is an important cost factor.
This approach in used in the PCI bus.

At any given time, one device is the bus master. It has the right to initiate data transfers
by issuing read and write commands. A master is called an initiator in PCI terminology. This is
other a processor or a DMA controller. The addressed device that responds to read and write
commands is called a target.
om
Device Configuration:-
When an I/O device is connected to a computer, several actions are needed to configure
both the device and the software that communicates with it.
c
p.
The PCI simplifies this process by incorporating in each I/O device interface a small
configuration ROM memory that stores information about that device. The configuration ROMs
ou
of all devices is accessible in the configuration address space. The PCI initialization software
reads these ROMs whenever the s ystem is powered up or reset. In each case, it determines
whether the device is a printer, a keyboard, an Ethernet interface, or a disk controller. It can
gr
further learn bout various device options and characteristics.
Devices are assigned addresses during the initialization process. This means that during
ts
the bus configuration operation, devices cannot be accessed based on their address, as they have
not yet been assigned one. Hence, the configuration address space uses a different mechanism.
en
Each device has an input signal called Initialization Device Select, IDSEL#.
The PCI bus has gained great popularity in the PC word. It is also used in many other
ud
computers, such as SUNs, to benefit from the wide range of I/O devices for which a PCI
interface is available. In the case of some processors, such as the Compaq Alpha, the PCI-
processor bridge circuit is built on the processor chip itself, further simplifying s ystem design
st
and packaging.
ity
3. Explain briefly bus arbitration phase in SCSI bus.(8M) June 2014,Jan 2014
.c
Ans. SCSI Bus:- The processor sends a command to the SCSI controller, which causes the
following sequence of event to take place:
w
1. The SCSI controller, acting as an initiator, contends for control of the bus.
w
2. When the initiator wins the arbitration process, it selects the target controller and hands
over control of the bus to it.
w
3. The target starts an output operation (from initiator to target); in response to this, the
initiator sends a command specifying the required read operation.

4. The target, realizing that it first needs to perform a disk seek operation, sends a message
to the initiator indicating that it will temporarily suspend the connection between them.
Then it releases the bus.
5. The target controller sends a command to the disk drive to move the read head to the first
sector involved in the requested read operation. Then, it reads the data stored in that
sector and stores them in a data buffer. When it is ready to begin transferring data to the
om
initiator, the target requests control of the bus. After it wins arbitration, it reselects the
initiator controller, thus restoring the suspended connection.
6. The target transfers the contents of the data buffer to the initiator and then suspends the
connection again. Data are transferred either 8 or 16 bits in parallel, depending on the
c
width of the bus.
p.
7. The target controller sends a command to the disk drive to perform another seek
operation. Then, it transfers the contents of the second disk sector to the initiator as
ou
before. At the end of this transfers, the logical connection between the two controllers is
terminated.
8. As the initiator controller receives the data, it stores them into the main memory using the
gr
DMA approach.
9. The SCSI controller sends as interrupt to the processor to inform it that the requested
ts
operation has been completed
en
This scenario show that the messages exchanged over the SCSI bus are at a higher
level than those exchanged over the processor bus. In this context, a higher level means that
ud
the messages refer to operations that may require several steps to complete, depending on the
device. Neither the processor nor the SCSI controller need be aware of the details of operation of
the particular device involved in a data transfer. In the preceding example, the processor need not
st
be involved in the disk seek operation.

ity
4.In a computer system why a PCI bus is used? With a neat sketch, explain how the read
operation is performed along with the role of IRDY#/TRDY# on the PCI bus (8M) Dec
.c
2014 ,JAN 2014.
Ans. Peripheral Component Interconnect (PCI) Bus:-

w
The PCI bus is a good example of a s ystem bus that grew out of the need for
w
standardization. It supports the functions found on a processor bus bit in a standardized format
that is independent of any particular processor. Devices connected to the PCI bus appear to the
w
processor as if they were connected directly to the processor bus. An important feature that the
PCI pioneered is a plug-and-play capability for connecting I/O devices. To connect a new device,
the user simply connects the device interface board to the bus. The software takes care of the
rest.

Data Transfer:-
In todays computers, most memory transfers involve a burst of data rather than just one
word. The reason is that modern processors include a cache memor y. Data are transferred
between the cache and the main memory in burst of several words each. The words involved in
om
such a transfer are stored at successive memory locations. When the processor (actually the
cache controller) specifies an address and requests a read operation from the main memory, the
memory responds by sending a sequence of data words starting at that address. Similarly, during
c
a write operation, the processor sends a memory address followed by a sequence of data words,
to be written in successive memory locations starting at the address. The PCI is designed
p.
primarily to support this mode of operation. A read or write operation involving a single word is
simply treated as a burst of length one.
ou
The bus supports three independent address spaces: memory, I/O, and configuration. The
first two are self-explanatory. The I/O address space is intended for use with processors, such as
gr
Pentium, that have a separate I/O address space. However, as noted , the system designer may
choose to use memory-mapped I/O even when a separate I/O address space is available. In fact,
ts
this is the approach recommended by the PCI its plug-and-play capability. A 4-bit command that
accompanies the address identifies which of the three spaces is being used in a given data
en
transfer operation.
ud
5.Draw the block diagram of universal bus(USB)structure connected to the host computer
Briefly explain all fields of packets that are used for communication between a host and a
device connected to an USB port. (8M) June 2012/July 2013
st
UNIVERSAL SERIAL BUS (USB):-

ity
The USB supports two speeds of operation, called low-speed (1.5 megabits/s) and full-
speed (12 megabits/s). The most recent revision of the bus specification (USB 2.0) introduced a
.c
third speed of operation, called high-speed (480 megabits/s). The USB is quickly gaining
acceptance in the market place, and with the addition of the high-speed capability it may well
w
become the interconnection method of choice for most computer devices.

w
The USB has been designed to meet several key objectives:

w
Provides a simple, low-cost and easy to use interconnection s ystem that overcomes the
difficulties due to the limited number of I/O ports available on a computer.
Accommodate a wide range of data transfer characteristics for I/O devices, including
telephone and Internet connections.
Enhance user convenience through a plug-and-play mode of operation

Port Limitation:-
The parallel and serial ports described in previous section provide a general-purpose
point of connection through which a variety of low-to medium-speed devices can be connected
to a computer. For practical reasons, only a few such ports are provided in a typical computer.
om
Device Characteristics:-
The kinds of devices that may be connected to a computer cover a wide range of
c
functionality. The speed, volume, and timing constraints associated with data transfers to and
p.
from such devices vary significantly.
A variety of simple devices that may be attached to a computer generate data of a similar
ou
nature low speed and asynchronous. Computer mice and the controls and manipulators used
with video games are good examples.
gr
Plug-and-Play:- As computers become part of everyday life, their existence should
become increasingly transparent. For example, when operating a home theater s ystem, which
ts
includes at least one computer, the user should not find it necessary to turn the computer off or to
restart the system to connect or disconnect a device.
en
The plug-and-play feature means that a new device, such as an additional speaker, can be
connected at any time while the system is operating. The s ystem should detect the existence of
ud
this new device automatically, identify the appropriate device-driver software and any other
facilities needed to service that device, and establish the appropriate addresses and logical
connections to enable them to communicate. The plug-and-play requirement has many
st
implications at all levels in the system, from the hardware to the operating s ystem and the
applications software. One of the primary objectives of the design of the USB has been to
ity
provide a plug-and-play capability.
The signaling convention on the PCI bus is similar to the one used, we assumed that the
.c
master maintains the address information on the bus until data transfer is completed. But, this is
not necessary. The address is needed only long enough for the slave to be selected. The slave can
w
store the address in its internal buffer. Thus, the address is needed on the bus for one clock cycle
only, freeing the address lines to be used for sending data in subsequent clock cycles. The result
w
is a significant cost reduction because the number of wires on a bus is an important cost factor.
This approach in used in the PCI bus.
w
At any given time, one device is the bus master. It has the right to initiate data transfers
by issuing read and write commands. A master is called an initiator in PCI terminology. This is
either a processor or a DMA controller. The addressed device that responds to read and write
commands is called a target.

6.a. Draw the hardware components needed for connecting a keyboard to a process and
explain in brief. Jan 2014
ANS: The CPU (Central Processing Unit) is the 'brain' of the computer.
It's typically a square ceramic package plugged into the motherboard, with a large heat sink on
om
top (and often a fan on top of that heat sink).
All instructions the computer will process are processed by the CPU. There are many "CPU
c
architectures", each of which has its own characteristics and trade-offs. The dominant CPU
p.
architectures used in personal computing are x86 and PowerPC. x86 is easily the most popular
processor for this class of machine (the dominant manufacturers of x86 CPUs
ou
are Intel and AMD). The other architectures are used, for istance, in workstations, servers or
embedded systems CPUs contain a small amount of static RAM (SRAM) called a cache. Some
gr
processors have two or three levels of cache, containing as much as several megabytes of
memory. ts
Dual Core
en
Some of the new processors made by Intel and AMD are Dual core. The Intel designation for
dual core are "Pentium D", "Core Duo" and "Core 2 Duo" while AMD has its "X2" series and
ud
"FX-6x".
The core is where the data is processed and turned into commands directed at the rest of the
st
computer. Having two cores increases the data flow into the processor and the command flow
out of the processor potentially doubling the processing power, but the increased performance is
ity
only visible with multithreaded applications and heavy multitasking.

.c
Motherboard
w
The Motherboard (also called Mainboard) is a large, thin, flat, rectangular fiberglass board
(typically green) attached to the case. The Motherboard carries the CPU, the RAM,
w
the chipset and theexpansion slots (PCI, AGP - for graphics -, ISA, etc.).
w
The Motherboard also holds things like the BIOS (Basic Input Output System) and the CMOS
Battery (a coin cell that keeps an embbeded RAM in the motherboard -often NVRAM- powered
to keep various settings in effect).

RAM
Random Access Memory (RAM) is a memory that the microprocessor uses to store data during
processing. This memory is volatile (loses its contents at power-down). When a software
application is launched, the executable program is loaded from hard drive to the RAM. The
om
microprocessor supplies address into the RAM to read instructions and data from it.
.b. List the SCSI bus signals with their functionalities. Jan 2014
c
Parallel SCSI (formally, SCSI Parallel Interface, or SPI) is one of the interface
p.
implementations in the SCSI family. In addition to being a data bus, SPI is a parallel electrical
bus: There is one set of electrical connections stretching from one end of the SCSI bus to the
ou
other. A SCSI device attaches to the bus but does not interrupt it. Both ends of the bus must
be terminated.
gr
ts
en
ud
st
ity
.c
w
w
w

UNIT-V- MEMORY SYSTEM

1. Draw the organization of a 1Kx1 memory cell and explain its working. (8M) June
2013,JAN 2014.
om
Ans. Another type of organization for 1k x 1 format is shown below:
5 Bit Row Address
W0
c
p.
5-Bit W132X32
ou
Decoder memory cell
gr
ts
10
en
Two 32 to 1
B W31
Multiplexers
I
ud
T Sence/Write
A Circuitr
st
D
ity
D
.c
The 10-bit address is divided into two groups of 5 bits each to form the row and column
w
addresses for the cell array. A row address selects a row of 32 cells, all of which are acc essed in
parallel. One of these, selected by the column address, is connected to the external data lines by
w
the input and output multiplexers. This structure can store 1024 bits, can be implemented in a 16-
pin chip.
w

2. Explain the working of a single-transistor dynamic memory cell. (10M) Dec 2014
Ans. Dynamic Memories:-
The basic idea of dynamic memory is that information is stored in the form of a charge
on the capacitor. An example of a dynamic memory cell is shown below:
om
When the transistor T is turned on and an appropriate voltage is applied to the bit line,
information is stored in the cell, in the form of a known amount of charge stored on the
capacitor. After the transistor is turned off, the capacitor begins to discharge. This is caused b y
c
the capacitors own leakage resistance and the very small amount of current that still flows
p.
through the transistor. Hence the data is read correctly only if is read before the charge on the
capacitor drops below some threshold value. During a Read
ou
gr
ts
operation, the bit line is placed in a high-impendance state, the transistor is turned on and a sense
en
circuit connected to the bit line is used to determine whether the charge on the capacitor is above
or below the threshold value. During such a Read, the charge on the capacitor is restored to its
original value and thus the cell is refreshed with every read operation.
ud
3. Define memory latency and bandwidth in case of burst operation that is used for
transferring a block of data to or from synchronous DRAM memory unit (8M) June 2012
st
Ans. Static Memories Vs Dynamic Memories:-

ity
Bipolar as well as MOS memory cells using a flip-flop like structure to store information
can maintain the information as long as current flow to the cell is maintained. Such memories are
.c
called static memories. In contracts, Dynamic memories require not only the maintaining of a
power suppl y, but also a periodic refresh to maintain the information stored in them. Dynamic
w
memories can have very high bit densities and very lower power consumption relative to static
memories and are thus generally used to realize the main memory unit.
w
w

4. Ex p l a i n any tw o cache ma p p in g f u n c t ions. J u l y 2 013 ,J a n 2 0 14

Ans: 1)DIRECT MAPPING
Each block of main memory maps to only one cache line
i.e. if a block is in cache, it must be in one specific place
Address is in two parts
Least Significant w bits identify unique word
om
Most Significant s bits specify one memory block
The MSBs are split into a cache line field r and a tag of s-r
2)ASSOCIATIVE MAPPING
Instead of placing main memory blocks in specific cache locations based on main
c
memory address, we could allow a block to go anywhere in cache.
p.
In this way, cache would have to fill up before any blocks move out.
This is how associative mapping cache works
ou
3)SET ASSOCIATIVE MAPPING
The problem of the direct mapping is eased by having a few choices for block placement.
At the same time, the hardware cost is reduced by decreasing the size of the associative
gr
mapping search.
Set associative mapping is a compromise that exhibits both the direct mapping and
ts
associative mapping while reducing their disadvantages
en
5. W i t h figur e e x p l ain abou t d i re c t ma p p i n g cache memor y . J an 2 0 1 4
Introduction
ud
Computer pioneers correctly predicted that programmers would want unlimited amounts of fast
memory. An economical solution to that desire is a memory hierarchy, which takes advantage of
st
locality and trade-offs in the cost-performance of memory technologies. The principle of locality,
presented in the first chapter, says that most programs do not access all code or data uniforml y.
ity
Locality occurs in time (temporal locality) and in space (spatial locality). This principle, plus the
guideline that for a given implementation technology and power budget smaller hardware can be
.c
made faster, led to hierarchies based on memories of different speeds and sizes. Figure 2.1 shows
w
a multilevel memory hierarchy, including typical sizes and speeds of access.

w
Since fast memory is expensive, a memory hierarchy is organized into several levels - each
w
smaller, faster, and more expensive per byte than the next lower level, which is farther from the
processor. The goal is to provide a memory s ystem with cost per byte almost as low as the
cheapest level of memory and speed almost as fast as the fastest level. In most cases (but not all),

the data contained in a lower level are a superset of the next higher level. This property, called
the inclusion property, is always required for the lowest level of the hierarchy, which consists of
main memory in the case of caches and disk memory in the case of virtual memory.
om
UNIT-VI: ARITHMETIC
1. Show the organization of virtual memory address translation based in fixed-length
c
pages and explains its working. (10M) June 2014
p.
Ans. Memory management by paging:- Fig 3 shows a simplified mechanism for virtual
ou
address translation in a paged MMU. The process begins in a manner similar to the segmentation
process. The virtual address composed of a high order page number and a low order word
number is applied to MMU. The virtual page number is limit checked to be certain that the page
gr
is within the page table, and if it is, it is added to the page table base to yield the page table entry.
The page table entry contains several control fields in addition to the page field. The control
ts
fields may include access control bits, a presence bit, a dirty bit and one or more use bits,
typically the access control field will include bits specifying read, write and perhaps execute
en
permission. The presence bit indicates whether the page is currently in main memory. The use bit
is set upon a read or write to the specified page, as an indication to the replaced algorithm in cas e
a page must be replaced.
ud
st
ity
.c
w
w
w
If the presence bit indicates a hit, then the page field of the page table entry will contain
the physical page number. If the presence bit is a miss, which is page fault, then the page field of

the page table entry which contains an address is secondary memory where the page is stored.
This miss condition also generates an interrupt. The interrupt service routine will initiate the
page fetch from secondary memory and with also suspended the requesting process until the
page has been bought into main memory. If the CPU operation is a write hit, then the dirty bit is
set. If the CPU operation is a write miss, then the MMU with begin a write allocate process.
om
2. Explain the design of a 4-bit carry-look a head adder. (8M) Dec 2014,Jan 2014
A n s.
c
p.
Carry-Look ahead Addition:
As it is clear from the previous discussion that a parallel adder is considerably slow & a
ou
fast adder circuit must speed up the generation of the carry signals, it is necessary to make the
carry input to each stage readily available along with the input bits. This can be achieved either
by propagating the previous carry or by generating a carry depending on the input bits &
gr
previous carry. The logic expressions for si (sum) and c i+1 (carry-out) of stage ith are
ts
en
ud
st
ity
.c
w
The above expressions Gi and Pi are called carry generate and propagate functions for
w
stage i. If the generate function for stage i is equal to 1, then ci+1 = 1, independent of the input
carry, ci. This occurs when both xi and yi are 1. The propagate function means that an input carry
w
will produce an output carry when either x i or yi or both equal to 1. Now, using Gi & Pi
functions we can decide carry for ith stage even before its previous stages have completed their
addition operations. All Gi and Pi functions can be formed independently and in parallel in only
one gate delay after the Xi and Yi inputs are applied to an n-bit adder. Each bit stage contains an

AND gate to form Gi, an OR gate to form Pi and a three-input XOR gate to form si. However, a
much simpler circuit can be derived by considering the propagate function as P i = xi yi, which
differs from Pi = xi + yi only when xi = yi =1 where Gi = 1 (so it does not matter whether Pi is 0
or 1). Then, the basic diagram in Figure-5 can be used in each bit stage to predict carry ahead of
any stage completing its addition.
om
Consider the ci+1expression,
c
This is because, Ci = (Gi-1 + Pi-1Ci-1).
p.
Further, Ci-1 = (Gi-2 + Pi-2Ci-2) and so on. Expanding in this fashion, the final carry expression
ou
can be written as below;
C i+1 = Gi + PiG i-1 + PiP i-1 G i-2 + + Pi P i-1 P 1G0 + Pi P i-1 P0G0
gr
Thus, all carries can be obtained in three gate delays after the input signals Xi, Yi and Cin
are applied at the inputs. This is because only one gate delay is needed to develop all Pi and Gi
ts
signals, followed by two gate delays in the AND-OR circuit (SOP expression) for ci + 1. After a
further XOR gate delay, all sum bits are available.
en
Therefore, independent of n, the number of stages, the As it is clear from the previous
discussion that a parallel adder is considerably slow & a fast adder circuit must speed up the
ud
generation of the carry signals, it is necessary to make the carry input to each stage readil y
available along with the input bits. This can be achieved either by propagating the previous carry
or by generating a carry depending on the input bits & previous carry. The logic ex pressions for
st
si (sum) and c i+1 (carry-out) of stage ith are

ity
.c
w
w
w

c om
p.
ou
gr
ts
en
FIG-5: 4 bit carry look ahead adder.

ud
st
Now, consider the design of a 4-bit parallel adder. The carries can be implemented as
;i = 0
ity
;i = 1
.c
;i = 2
w
;i = 3
w
The complete 4-bit adder is shown in Figure 5b where the B cell indicates Gi, Pi & Si
generator. The carries are implemented in the block labeled carry look-ahead logic. An adder
w
implemented in this form is called a carry look ahead adder. Delay through the adder is 3 gate
delays for all carry bits and 4 gate dela ys for all sum bits. In comparison, note that a 4-bit ripple-
carry adder requires 7 gate delays for S3(2n-1) and 8 gate delays(2n) for c4.
3.Answer the following w.r.t. to magnetic disk,the secondary storage device: (6M)

June2012
i)seek time
ii) latency
iii) access time
om
1. Ans. Seek time: - Is the average time required to move the read/write head to the desired
track. Actual seek time which depend on where the head is when the request is received
and how far it has to travel, but since there is no way to know what these values will be
c
when an access request is made, the average figure is used. Average seek time must be
p.
determined by measurement. It will depend on the physical size of the drive components
and how fast the heads can be accelerated and decelerated. Seek times are generally in the
ou
range of 8-20 m sec and have not changed much in recent years.
2. Track to track access time: - Is the time required to move the head from one track to
adjoining one. This time is in the range of 1-2 m sec.
gr
3. Rotational latency: - Is the average time required for the needed sector to pass under
head once and head has been positioned once at the correct track. Since on the average
the desired sector will be half way around the track from where the head is when the head
ts
first arrives at the track, rotational latency is taken to be the rotation time. Current
rotation speeds are from 3600 to 7200 rpm, which yield rotational latencies in the 4-8 ms
en
range.
4. Average Access time:- Is equal to seek time plus rotational latency.
ud
5. With f i g ure exp l a i n c i r cuit arrange m e nts f o r binary div i s ion J a n 2014
A circuit that implements division by this longhand method operates as follows: It
st
positions the divisor appropriately with respect to the dividend and performs a subtraction.
If the remainder is zero positive, a quotient bit of 1 is determined, the remainder isextended
ity
by another bit of the dividend, the divisor is repositioned, and subtraction is performed.
On the other hand, if the remainder is negative, a quotient bit of 0 is determined,
.c
the dividend is restored by adding back the divisor, and the divisor H
w
w
w

c om
p.
ou
gr
ts
en
ud
6. I E EE standard f o r f l o a ting p o i nt numbers,ex p l a i n . J a n 2014

st
The standard defines

ity
arithmetic formats: sets of binary and decimal floating-point data, which consist of finite
.c
numbers (including signed zeros and subnormal numbers), infinities, and special "not a
number" values (NaNs)
w
interchange formats: encodings (bit strings) that may be used to exchange floating-point data
w
in an efficient and compact form

rounding rules: properties to be satisfied when rounding numbers during arithmetic and
w
conversions
operations: arithmetic and other operations on arithmetic formats

exception handling: indications of exceptional conditions (such as division by zero,

overflow,etc.)
The standard also includes extensive recommendations for advanced exception handling,
additional operations (such as trigonometric functions), expression evaluation, and for achieving
om
reproducible results.
The standard is derived from and replaces IEEE 754-1985, the previous version, following a
seven-year revision process, chaired by Dan Zuras and edited by Mike Cowlishaw. The binary
c
formats in the original standard are included in the new standard along with three new basic
p.
formats (one binary and two decimal). To conform to the current standard, an implementation
must implement at least one of the basic formats as both an arithmetic format and an interchange
ou
format.
gr
ts
en
ud
st
ity
.c
w
w
w

UNIT-VII: BASIC PROCESSING UNIT

1. Performance signed multiplication of numbers 13 and -6 using both multiplication
algorithm represent the numbers in 5-bit including sign bit .Give booth multiplier
recording table that is used in the above multiplication (10M) Dec 2013
om
Booth Algorithm
The Booth algorithm generates a 2n-bit product and both positive and negative 2's-
c
complement n-bit operands are uniformly treated. To understand this algorithm, consider a
p.
multiplication operation in which the multiplier is positive and has a single block of 1s, for
example, 0011110(+30). To derive the product, as in the normal standard procedure, we could
ou
add four appropriately shifted versions of the multiplicand,. However, using the Booth algorithm,
we can reduce the number of required operations by regarding this multiplier as the difference
between numbers 32 & 2 as shown below;
gr
ts
0 1 0 0 0 0 0 (32)
en
0 0 0 0 0 1 0 (-2)
ud
This suggests that the product can be generated by adding 25 times the multiplicand to the
st
2's-complement of 21 times the multiplicand. For convenience, we can describe the sequence of
required operations by recoding the preceding multiplier as 0 +1000 - 10. In general, in the
ity
Booth scheme, -1 times the shifted multiplicand is selected when moving from 0 to 1, and +1
times the shifted multiplicand is selected when moving from 1 to 0, as the multiplier is scanned
from right to left.
.c
w
w
w

om
0 1 0 1 1 0 1
0 0+1 +1 +1 +1 0
c
0 0 0 0 0 0 0
p.
0 1 0 1 1 0 1
ou
0 N1or0ma1l M
1 ul0tipli1cation
0 1 0 01 11 00 11 1 0 1
gr
0 0+1 +1 +1 +1 0
ts
0 0 0 0 0 0 00 0 0 0 0 0 0
en
1 1 1 1 1 1 1 01 0 0 1 1
ud
0 0 0 0 0 0 0 00 0 0 0
0 0 0 0 0 0 0 00 0 0
st
ity
FIG-10b: Booth Multiplication
Figure 10 illustrates the normal and the Booth algorithms for the said example. The
.c
Booth algorithm clearly extends to any number of blocks of 1s in a multiplier, including the
situation in which a single 1 is considered a block. See Figure 11a for another example of
w
recoding a multiplier. The case when the least significant bit of the multiplier is 1 is handled by
assuming that an implied 0 lies to its right. The Booth algorithm can also be used directly for
w
negative multipliers, as shown in Figure 11a.

w
To verify the correctness of the Booth algorithm for negative multipliers, we use the following
property of negative-number representations in the 2's-complement

c om
p.
ou
gr
ts
en
ud
2. Perform division of numbers 8 by 3 using non-restoring division algorithm(8M)June2014
Ans. No restoring Division:

st
The restoring-division algorithm can be improved by avoiding the need for restoring A after an
unsuccessful subtraction. Subtraction is said to be unsuccessful if the resultis negative. Consider
ity
the sequence of operations that takes place after the subtraction operation in the preceding
algorithm. If A is positive, we shift left and subtract M, that is, we perform 2A - M. If A is
.c
negative, we restore it by performing A + M, and then we shift it left and subtract M. This is
equivalent to performing 2A + M. The q0 bit is appropriately set to 0 or 1 after the correct
w
operation has been performed. We can summarize this in the following algorithm for no
restoring division.
w
Step 1: Do the following n times:

w
1.If the sign of A is 0, shift A and Q left one bit position and subtract M fromA; otherwise, shift
A and Q left and add M to A.
2. Now, if the sign of A is 0, set q0 to 1; otherwise, set q0 to 0.

Step 2: If the sign of A is 1, add M to A.
Step 2 is needed to leave the proper positive remainder in A at the end of the n cycles of
Step 1. The logic circuitry in Figure 17 can also be used to perform this algorithm. Note that the
Restore operations are no longer needed, and that exactly one Add or Subtract operation is
om
performed per cycle. Figure 19 shows how the division example in Figure 18 is executed by the
no restoring-division algorithm.
There are no simple algorithms for directly performing division on signed operands that are
c
comparable to the algorithms for signed multiplication. In division, the operands can be
p.
preprocessed to transform them into positive values. After using one of the algorithms just
discussed, the results are transformed to the correct signed values, as necessary.
ou
gr
ts
en
ud
st
ity
.c
w
FIG 19: Non-restoring Division

w
w

3.Write the algorithm for binary division using restoring division method(8M)
Dec2014/July 2013
Ans. Restoring Division:
Figure 17 shows a logic circuit arrangement that implements restoring division. Note its
similarity to the structure for multiplication that was shown in Figure 8. An n-bit positive divisor
om
is loaded into register M and an n-bit positive dividend is loaded into register Q at the start of the
operation. Register A is set to 0. After the division is complete, the n-bit quotient is in register Q
and the remainder is in register A. The required subtractions are facilitated by using 2's-
c
complement arithmetic. The extra bit position at the left end of both A and M accommodates the
sign bit during subtractions. The following algorithm performs restoring division.
p.
Do the following n times:
ou
1. Shift A and Q left one binary position.
gr
2. Subtract M from A, and place the answer back in A.
3. If the sign of A is 1, set q0 to 0 and add M back to A (that is, restore A); otherwise, set
ts
q0to 1.
Figure 18 shows a 4-bit example as it would be processed by the circuit.
en
ud
4.. List the rules for addition, subtraction, multiplication and d ivision of floating point
numbers (8M) June 2012/July 2013
Ans. FLOATING-POINT NUMBERS AND OPERATIONS:

st
ity
Floating point arithmetic is an automatic way to keep track of the radix point. The
discussion so far was exclusively with fixed-point numbers which are considered as integers, that
.c
is, as having an implied binary point at the right end of the number. It is also possible to assume
that the binary point is just to the right of the sign bit, thus representing a fraction or any where
w
else resulting in real numbers. In the 2's-complement system, the signed value F, represented by
the n-bit binary fraction
w
B = b0.b - 1b -2 ..b-(n-1) is given by

w
F(B) = -bo x 2 + b-1 x 2-1 +b-2x2-2 + ... + b-(n-X) x 2-{n~l) where the range of F is
-1 F 1 -2-(n-1). Consider the range of values representable in a 32-bit, signed, fixed-point

format. Interpreted as integers, the value range is approximately 0 to 2.15 x 109. If we consider
them to be fractions, the range is approximately 4.55 x 10-10 to 1. Neither of these ranges is

sufficient for scientific calculations, which might involve parameters like Avogadro's number
(6.0247 x 1023 mole-1) or Planck's constant (6.6254 x 10-27erg s). Hence, we need to easil y
accommodate both very large integers and very small fractions. To do this, a computer must be
able to represent numbers and operate on them in such a way that the position of the binary point
is variable and is automatically adjusted as computation proceeds. In such a case, the binary
point is said to float, and the numbers are called floating-point numbers. This distinguishes them
om
from fixed-point numbers, whose binary point is always in the same position.
Because the position of the binary point in a floating-point number is variable, it must be
given explicitly in the floating-point representation. For example, in the familiar decimal
c
scientific notation, numbers may be written as 6.0247 x 1023, 6.6254 -10-27, -1.0341 x 102, -
p.
7.3000 x 10-14, and so on. These numbers are said to be given to five significant digits. The scale
factors (1023, 10-27, and so on) indicate the position of the decimal point with respect to the
ou
significant digits. By convention, when the decimal point is placed to the right of the first
(nonzero) significant digit, the number is said to be normalized. Note that the base, 10, in the
scale factor is fixed and does not need to appear explicitly in the machine representation of a
gr
floating-point number. The sign, the significant digits, and the exponent in the scale factor
constitute the representation. We are thus motivated to define a floating-point number
ts
representation as one in which a number is represented by its sign, a string of significant digits,
commonly called the mantissa, and an exponent to an implied base for the scale factor.
en
4. Draw a n d e x p l ain the s i n gle-b u s o r g anization of tha data path i n s i de a
processor. J a n 2014
ud
Ans: A datapath is a collection of functional units, such as arithmetic logic units or multipliers,
that perform data processing operations. It is a central part of many central processing
st
units (CPUs) along with the control unit, which largely regulates interaction between the
datapath and the data itself, usually stored in registers or main memory.
ity
Recently, there has been growing research in the area of reconfigurable datapathsdatapaths
that may be re-purposed at run-time using programmable fabricas such designs may allow for
.c
more efficient processing as well as substantial power savings

w
Examples
w
Let us consider addition as an Arithmetic operation and Retrieving data from memory in detail.
w
Example 1) Arithmetic addition : contents of register reg1 and reg2 are added and the result is
stored in reg3
Sequence of operations:

1. reg1out,Xin
2. reg2out,choose X,ADDITION,Yin
3. Yout,reg3in
The control signals written in one line are executed in the same clock cycle. all other signals
om
remain untouched. So, in the first step the contents of register1 are written into the register X
through the bus. In the second stage the content of register2 is placed onto the bus and
theMultiplexer is made to choose input X as the contents of reg1 are stored in register X. The
c
ALU then adds the contents in the register X and reg2 and stores the result of the addition in the
p.
special temporary register Y. In the final step the result strored in Y is sent over to the register
reg3 over the internal processor bus. Only one register can output its data onto bus in one step
ou
4. control sequence for an un-co n d itional branch i n struct i o n J a n 2 0 14
gr
Ans: A branch is an instruction in a computer program that may, when executed by a computer,
cause the computer to begin execution of a different instruction
ts
sequence. Branch (or branching,branched) may also refer to the act of beginning execution of a
different instruction sequence due to executing a branch instruction. A branch instruction can be
en
either an unconditional branch, which always results in branching, or a conditional branch,
which may or may not cause branching depending on some condition.
ud
When executing (or "running") a program, a computer will fetch and execute instructions in
sequence (in their order of appearance in the program) until it encounters a branch instruction. If
st
the instruction is an unconditional branch, or it is conditional and the condition is satisfied, the
ity
computer will branch (fetch its next instruction from a different instruction sequence) asspecified
by the branch instruction. However, if the branch instruction is conditional and the condition is
.c
not satisfied, the computer will not branch; instead, it will continue executing the current
instruction sequence, beginning with the instruction that follows the conditional branch
w
instruction.
w
w

UNIT-VIII: MULTICORES, MULTIPROCESSORS, AND CLUSTERS
1. Write and explain the control sequences for the execution of an unconditional branch
om
instruction(10M) June 2014
Ans. Branch Instructions:
c
A branch instruction replaces the contents of the PC with the branch target address. This
address is usually obtained by adding an offset X, which is given in the branch instruction, to the
p.
updated value of the PC. Listing in figure 8 below gives a control sequence that implements an
unconditional branch instruction. Processing starts, as usual, with the fetch phase. This phase
ou
ends when the instruction is loaded into the IR in step 3. The offset value is extracted from the IR
by the instruction decoding circuit, which will also perform sign extension if required. Since the
value of the updated PC is already available in register Y, the offset X is gated onto the bus in
gr
step 4, and an addition operation is performed. The result, which is the branch target address, is
loaded into the PC in step 5. ts
The offset X used in a branch instruction is usually the difference between the branch target
en
address and the address immediately following the branch instruction.
ud
st
ity
.c
]
w
w
w
For example, if the branch instruction is at location 2000 and if the branch target address
is 2050, the value of X must be 46. The reason for this can be readily appreciated from the
control sequence in Figure 7. The PC is incremented during the fetch phase, before knowing the
type of instruction being executed. Thus, when the branch address is computed in step 4, the PC

value used is the updated value, which points to the instruction following the branch instruction
in the memory.
Consider now a conditional branch. In this case, we need to check the status of the condition
codes before loading a new value into the PC. For example, for a Branch-on-negative
(Branch<0) instruction, step 4 is replaced with
om
2.Explain with block diagram the basic organization of a micro programmed control
unit(10M) Dec 2013
c
p.
Ans. MICROPROGRAMMED CONTROL:ALU is the heart of any computing s ystem, while
Control unit is its brain. The design of a control unit is not unique; it varies from designer to
ou
designer. Some of the commonly used control logic design methods are;
Sequence Reg & Decoder method

Hard-wired control method
gr
PLA control method
Micro-program control method
ts
en
ud
st
ity
.c
w
The control signals required inside the processor can be generated using a control step
w
counter and a decoder/ encoder circuit. Now we discuss an alternative scheme, called micro
programmed control, in which control signals are generated by a program similar to machine
w
language programs.

c om
p.
ou
First, we introduce some common terms. A control word (CW) is a word whose
individual bits represent the various control signals in Figure 12. Each of the control steps in the
control sequence of an instruction defines a unique combination of Is and Os in the CW. The
gr
CWs corresponding to the 7 steps of Figure 6 are shown in Figure 15. We have assumed that
Select Y is represented by Select = 0 and Select4 by Select = 1. A sequence of CWs
ts
corresponding to the control sequence of a machine instruction constitutes the micro routine for
that instruction, and the individual control words in this micro routine are referred to as
en
microinstructions.
The micro routines for all instructions in the instruction set of a computer are stored in a
special memory called the control store. The control unit can generate the control signals for any
ud
instruction by sequentially reading the CWs of the corresponding micro routine from the control
store. This suggests organizing the control unit as shown in Figure 16. To read the control words
st
sequentially from the control store, a micro program counter (PC) is used. Every time a new
instruction is loaded into the IR, the output of the block labeled "starting address generator" is
ity
loaded into the PC. The PC is then automatically incremented by the clock, causing
successive microinstructions to be read from the control store. Hence, the control signals are
delivered to various parts of the processor in the correct sequence.
.c
One important function of the control unit cannot be implemented by the simple organization
w
in Figure 16. This is the situation that arises when the control unit is required to check the status
of the condition codes or external inputs to choose between alternative courses of action. In the
w
case of hardwired control, this situation is handled by including an appropriate logic function, in
the encoder circuitry. In micro programmed control, an alternative approach is to use conditional
w
branch microinstructions. In addition to the branch address, these microinstructions specify

which of the external inputs, condition codes, or, possibly, bits of the instruction register should
be checked as a condition for branching to take place.

3. What are modifications required in the basic organization of a microprogrammed

control unit to support conditional branching in the microprogram (8M) June 2012,Jan
201 4
Ans. However, this scheme has one serious drawback assigning individual bits to each
control signal results in long microinstructions because the number of required signals is usuall y
om
large. Moreover, only a few bits are set to 1 (to be used for active gating) in any given
microinstruction, which means the available bit space is poorly used. Consider again the simple
processor of Figure 2, and assume that it contains only four general-purpose registers, R0, Rl,
R2, and R3. Some of the connections in this processor are permanently enabled, such as the
c
output of the IR to the decoding circuits and both inputs to the ALU. The remaining connections
p.
to various registers require a total of 20 gating signals. Additional control signals not shown in
the figure are also needed, including the Read, Write, Select, WMFC, and End signals. Finally,
ou
we must specify the function to be performed by the ALU. Let us assume that 16 functions are
provided, including Add, Subtract, AND, and XOR. These functions depend on the particular
ALU used and do not necessarily have a one-to-one correspondence with the machine instruction
gr
OP codes. In total, 42 control signals are needed.
If we use the simple encoding scheme described earlier, 42 bits would be needed in each
ts
microinstruction. Fortunately, the length of the microinstructions can be reduced easily. Most
signals are not needed simultaneously, and many signals are mutually exclusive. For example,
en
only one function of the ALU can be activated at a time. The source for a data transfer must be
unique because it is not possible to gate the contents of two different registers onto the bus at the
same time. Read and Write signals to the memory cannot be active simultaneously. This suggests
ud
that signals can be grouped so that all mutually exclusive signals are placed in the same group.
Thus, at most one micro operation per group is specified in any microinstruction. Then it is
possible to use a binary coding scheme to represent the signals within a group. For example, four
st
bits suffice to represent the 16 available functions in the ALU. Register output control signals
can be placed in a group consisting of PCout, MDRout, Zout, Offsetout, R0out Rlout, R2out, R3out, and
ity
TEMPout. Any one of these can be selected by a unique 4-bit code.

.c
Further natural groupings can be made for the remaining signals. Figure 19 shows an
example of a partial format for the microinstructions, in which each group occupies a field large
w
enough to contain the required codes. Most fields must include one inactive code for the case in
which no action is required. For example, the all-zero pattern in Fl indicates that none of the
w
registers that may be specified in this field should have its contents placed on the bus. An
inactive code is not needed in all fields. For example, F4 contains 4 bits that specify one of the
w
16 operations performed in the ALU. Since no spare code is included, the ALU is active during
the execution of every microinstruction. However, its activity is monitored by the rest of the
machine through register Z, which is loaded only when the Zin signal is activated.

Grouping control signals into fields requires a little more hardware because decod ing
circuits must be used to decode the bit patterns of each field into individual control signals. The
cost of this additional hardware is more than offset by the reduced number of bits in each
microinstruction, which results in a smaller control store. In Figure 19, only 20 bits are needed to
store the patterns for the 42 signals.
om
So far we have considered grouping and encoding only mutually exclusive co ntrol
signals. We can extend this idea by enumerating the patterns of required signals in all possible
microinstructions. Each meaningful combination of active control signals can.
c
4. List out differences between shared memory multiprocessor and cluster July 2013,Jan
201 4
p.
Ans: In a multiple processor computer, an important issue is: How do processors coordinate to
ou
solve a problem? Processors must have the ability to communicate with each other in order to
cooperatively complete a task. There are two general approaches to address this problem.
gr
One option uses a single address space. Systems based on this concept, otherwise known as
shared-memory systems, allow processor communication through variables stored in a shared
address space.
ts
The other alternative employs a scheme by which each processor has its own memory module.
en
Such a distributed-memory system (cluster) is constructed by connecting each component with a
high-speed communications network. Processors communicate to each other over the network.
ud
The architectural differences between shared-memory systems and distributed-memory systems

have implications on how each is programmed. With a shared-memory multiprocessor, different
processors can access the same variables. This makes referencing data stored in memory similar
st
to traditional single-processor programs, but adds the complexity of shared data integrity. A
distributed-memory system introduces a different problem: how to distribute a computational
ity
task to multiple processors with distinct memory spaces and reassemble the results from each
processor into one solution.
.c
Distributed Computing is a way of combining the processing power of thousands of small

w
computers (ie: PCs) to solve very complex problems that are too large for traditional
supercomputers, which are very expensive to build and run.
w
5. What are the classifications of parallel processing Jan 2014

1.Parallel computing is a form ofcomputation in which many calculations are carried out
w
simultaneously, operating on the principle that large problems can often be divided into smaller
ones, which are then solved concurrently("in parallel"). There are several different forms of
parallel computing: bit-level,instruction level, data, and tasparallelism. Parallelism has been

employed for many years, mainly in high-performance computing, but interest in it has grown
lately due to the physical constraints preventing frequency scaling. As power consumption (and
consequently heat generation) by computers has become a concern in recent years,[3] parallel
computing has become the dominant paradigm in computer architecture, mainly in the form
of multicore processors.
om
2.Parallel computers can be roughly classified according to the level at which the
hardware supports parallelism, with multi-core and multi-processor computers having
c
multiple processing elements within a si n g l e machine, while clusters, MPPs,
p.
and grids use multiple computers to work on the same task. Specialized parallel computer
architectures are sometimes used alongside traditional processors, for accelerating
ou
specific tasks.
3.Parallel computer programs are more difficult to write than sequential ones, because
gr
concurrency introduces several new classes of potential software bugs, of which race
conditionsare the most common. Communication and synchronization between the
ts
different subtasks are t ypically some of the greatest obstacles to get ting good parallel
program performance.
en
ud
st
ity
.c
w
w
w

CO

Uploaded by

Copyright:

Available Formats

CO

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

CO

Uploaded by

Copyright:

Available Formats

COMPUTER ORGANIZATION 10CS46

Input/Output Organization: Accessing I/O Devices, Interrupts Interrupt Hardware,

Requests, Exceptions, Direct Memory Access, Buses

Input/Output Organization contd.: Interface Circuits, Standard I/O Interfaces PCI

Dept of CSE,SJBIT Page 1

from Uniprocessors to Multiprocessors, Amdahls Law, Shared Memory

Multithreading, SISD, IMD, SIMD, SPMD, and

Tata McGraw Hill, 2002. (Listed topics only from Chapters 1, 2, 4, 5, 6, 7)

2. David A. Patterson, John L. Hennessy: Computer Organization and Design The

Dept of CSE,SJBIT Page 2

Dept of CSE,SJBIT Page 3

Operational Concepts, Bus Structures, Performance Processor Clock,

Dept of CSE,SJBIT Page 4

1.1 Computer types

in engineering applications of interactive design work.

worldwide source of all types of information.

required in the applications like weather forecasting etc.,

1.2 Functional unit

A computer consists of five functionally independent main parts input, memory,

Dept of CSE,SJBIT Page 5

Joysticks, trackballs, mouse, scanners etc are other input devices.

Dept of CSE,SJBIT Page 6

To provide easy access to a word in memory, a distinct address is associated with

Arithmetic logic unit (ALU):-

Examples:- Printer, speakers, monitor etc.

Dept of CSE,SJBIT Page 7

2. The operand at LOCA is fetched and added to the contents of R0

Dept of CSE,SJBIT Page 8

Fig b : Connections between the processor and the memory

control the various processing elements in one execution of instruction.

The program counter PC:-

Dept of CSE,SJBIT Page 9

Normal execution of a program may be preempted (temporarily interrupted) if

The processor provides the requested service by executing an appropriate interrupt

Dept of CSE,SJBIT Page 10

Downloaded from www.citystudentsgroup.com

1.4 Bus structure

Fig c: Single bus structure

Single bus structure is

Very flexible for attaching peripheral devices

Dept of CSE,SJBIT Page 11

Downloaded from www.citystudentsgroup.com

Fig d: The processor cache

Dept of CSE,SJBIT Page 12

Downloaded from www.citystudentsgroup.com

1.6 Basic performance equation

Dept of CSE,SJBIT Page 13

Downloaded from www.citystudentsgroup.com

A higher degree of concurrency can be achieved if multiple instructions pipelines

with such an arrangement, it becomes possible to start the execution of several

Dept of CSE,SJBIT Page 14

Downloaded from www.citystudentsgroup.com

1.7 Clock rate

These are two possibilities for increasing the clock rate R.

instruction a large number of instructions may be needed to perform a given

But complex instructions combined with pipelining (effective value of S 1)

1.8 Performance measurements

It is very important to be able to access the performance of a computer, comp

The previous argument suggests that the performance of a computer is given by