UNIT-2

COMPUTER SYSTEM OVERVIEW

Lecture-16
Technology :

➢ Technology shapes the computer for better performance.

➢ Technologies that have been used over time with relative performance per unit cost for
each technology.

Year Technology used in computer Relative performance/unit cost

1951 Vacuum tube 1
1965 Transistor 35
1975 Integrated circuit 900
1995 Very large – scale IC’s 2400.00
2013 Ultra large-scale IC’s 6,200,000.00

Vacuum Tubes:

➢ The first electronic computer, ENIAC (Electronic Numerical Integrator and computer)
➢ It was designed and constructed by Eckert and Mauchly.
➢ It was made up of more than 18000 vacuum tubes and 1500 relays.
➢ It was able to perform nearly 5000 additions or subtractions per second.
➢ It was a decimal rather than a binary machine.
➢ Weight-30 tones, area -15000 sq.ft , power consumption -140kW.
➢ Data memory consists of 20 accumulators, each capable of storing a ten digit decimal
number.

Transistors:

➢ A transistor is simply an on/off switch controlled by electricity.

➢ Transistors are smaller, cheaper and low power consumption
➢ Greater speed, larger memory capacity and smaller size than first generation.
➢ CPU can handle both floating point and fixed point operation.
➢ Separate I/O processor having direct access to main memory to control I/O operations.
➢ It introduction of more complex arithmetic and logic unit and control units to support
high level languages.

Integrated circuit:

➢ It enabled lower cost, faster processors and development of memory chips.

➢ IC allowed to increase memory size and number of I/O port
➢ Magnetic core memories were replaced by integrated circuit memories.
➢ IC combined dozens to hundreds of transistors into a single chip.

VLSI: very large-scale integrated (VLSI) circuit

It consists of billions of combinations of conductors, insulators & switches

manufactured in a small package.
■ Excellent conductors of electricity (using either microscopic copper or aluminum wire)
■ Excellent insulators from electricity (like plastic sheathing or glass)
■ Areas that can conduct or insulate under special conditions (as a switch)

ULSI:

➢ Is the process of integrating or embedding millions of transistors on a single silicon

semiconductor micro chip.
➢ It is a successor to large scale integration and very large scale integrating technology
➢ It was design to provide the greatest possible computational power from the smallest
form factor of microchip or microprocessor dye.

Performance:

➢ Performance is an important attribute of a computer.

➢ It is an important criterion for selection of a computer.
➢ Performance of a computer can be measured in number of ways.

Performance based on:

1. Response time:
➢ How long it takes to do a task.
➢ It is also called execution time.
➢ It includes disk access, memory access, I/O activities.

2. Throughput :
➢ Total amount of work done in a given time.
➢ It is also called bandwidth.

Performance of computer is directly related to the throughput and hence it is reciprocal of

execution time.
1
𝑝𝑒𝑟𝑓𝑜𝑟𝑚𝑎𝑛𝑐𝑒 =
Executiontime

1
Executiontime B
ti
m
e
A

Execution time B > Execution A

Execution time B greater than execution A so A is faster than B

A is n times faster as B to mean

performance A
=n
performance B

performance A 1 1
= /
performance B Executiontime A Executiontime B
Problem:

If a computer A runs a program in 10 second & B runs the same problem in 15 seconds how
much faster is A than B?

performance A Execution time B

= =n
performance B Executiontime A

15
= 1.5
10
A is therefore 1.5 times as fast as B

performance A
= 1.5
performance B

performance A
= performance B
1.5
Measuring performance:

➢ One of the important measures of a computer performance is a time.

➢ Program execution time is measured in seconds per program.
➢ Time can be divided in different ways

CPU time:
➢ It is also called CPU execution time
➢ The actual time the cpu spends for computing a specific task.
➢ CPU time is divided into user cpu time and system cpu time.

User CPU time: The cpu time spent in a program

System cpu time: The cpu time spend in the operating system perform task on behalf of the program.

Performance metrics:

➢ Users and designers often examine performance using different metrics.

➢ All computers are constructed using a clock that determines when events take place in
the hardware

Most basic metrics are:

1. Clock cycle : A clock cycle, or simply a "cycle," is a single electronic pulse of a CPU.
During each cycle, a CPU can perform a basic operation such as fetching an instruction,
accessing memory, or writing data.
2. Clock cycle time :

𝐶𝑃𝑈 𝑒𝑥𝑒𝑐𝑢𝑡𝑖𝑜𝑛 𝑡𝑖𝑚𝑒 𝑓𝑜𝑟 𝑎 𝑝𝑟𝑜𝑔𝑟𝑎𝑚 = 𝑐𝑝𝑢 𝑐𝑙𝑜𝑐𝑘 𝑐𝑦𝑐𝑙𝑒 𝑓𝑜𝑟 𝑎 𝑝𝑟𝑜𝑔𝑟𝑎𝑚 𝑋 𝑐𝑙𝑜𝑐𝑘 𝑐𝑦𝑐𝑙𝑒 𝑡𝑖𝑚𝑒

Alternatively because clock rate and clock cycle time are inverses:

CPU clock cycles for a program

𝐶𝑃𝑈 𝑒𝑥𝑒𝑐𝑢𝑡𝑖𝑜𝑛 𝑡𝑖𝑚𝑒 𝑓𝑜𝑟 𝑎 𝑝𝑟𝑜𝑔𝑟𝑎𝑚 =
Clock rate

3. Clock rate:

The clock rate of a computer is normally determined bythe frequency of a crystal.

Instruction performance:

The computer had to execute the instructions to run the program.

The execution time must depend on the number of instructions in a program.

CPU clock cycles =instructions for a program * Average clock cycles per instructions.

Clock cycle per instructions (CPI):

Average number of clock cycles per instructions for a program or program fragment

CPU clock cycles

𝐶𝑃𝐼 =
Instruction count

The classic cpu performance equation:

Cpu time = instruction count * CPI * clock cycle time

Since the clock rate is the inverse of the clock cycle time.

instruction count ∗ CPI

cpu time =
clock rate

IC : the number of instructions executed by the program.

Describe the eight ideas that leads to performance improvement (8)
Eight ideas that the computer architecture has been invented for computer design.

1. Design for moore’s law

2. Use abstraction for design
3. Make common case fast
4. Performance via parallelism
5. Performance via pipelining
6. Performance via prediction
7. Hierarchy of memory
8. Dependability via redundancy

Design for moore’s law :

➢ It states that integrated circuit resources (transistors) double every 18–24 months.
➢ The computer designer must predict the rapid change in IC capacity & design it
accordingly.
➢ Moore’s Law graph to represent designing for rapid change.

Use abstraction for design:

➢ Abstraction means freedom from representational quality

➢ A major productivity technique for hardware and software is to use abstractions.
➢ Use abstractions to represent the design at different levels of representation.
➢ Lower-level details are hidden to offer a simpler model at higher levels.

Make the Common Case Fast :

➢ Make the common case fast to enhance performance better than optimizing the rare
case.
➢ For this idea the common case has to be carefully identified and experimented.
➢ Example: increasing speed level for a sports car is very easier than to a minivan

Performance via parallelism:

➢ Parallelism means simultaneous execution of source task on multiple processors in

order to obtain the result faster
➢ Computer architects have offered designs that get more performance by performing
operations in parallel.
Example: Dual- quad processor

Performance via pipelining :

➢ A particular pattern of parallelism is called pipelining.

➢ In pipelining more than one instruction are executed at the same time to increase the
performance and throughput.
Performance via prediction:

➢ A statement about what will happen or might happen in the future.

➢ In some cases, based on prediction.
➢ It is better to start working based on prediction or average guess to make the
performance faster rather than working until you know for sure.

Hierarchy of Memories :

➢ Memory speed and size often plays a vital role in increasing the performance of the system,
but due to the high cost of memory, the size of problem that can be solved is limited.
➢ To address this demand, hierarchy of memory has to be used.
➢ Memory to be faster, smallest and most expensive memory per bit at the top of the hierarchy
and the slowest, largest and cheapest memory bit at the bottom of the hierarchy.

size speed

Factors that impact the overall performance of a computer

1. CPU Performance:
• Clock Speed: The clock speed of the CPU is a crucial factor. It represents the number of
instructions a CPU can execute per second. Higher clock speeds generally result in better
performance.
• Number of Cores and Threads: Multi-core processors can handle multiple tasks
simultaneously, and hyper-threading allows each core to execute multiple tasks. This is particularly
beneficial for multitasking and parallel processing.
2. Memory (RAM) Performance:
• Capacity: The amount of RAM affects the system's ability to handle multiple applications
and large datasets. Inadequate RAM can lead to slow performance, especially when running
memory-intensive applications.
• Speed: The speed of the RAM (measured in MHz) influences how quickly data can be read
from and written to memory. Faster RAM can improve overall system responsiveness.
3. Storage Performance:
• Type of Storage (HDD vs. SSD): Solid State Drives (SSDs) are significantly faster than Hard
Disk Drives (HDDs) because they have no moving parts. Upgrading to an SSD can dramatically
improve system responsiveness and reduce loading times.
 • Read/Write Speeds: The read and write speeds of a storage device impact how quickly
data can be retrieved or saved. Faster speeds result in quicker application loading and improved
overall system performance.
4. Motherboard and Chipset:
• The motherboard and chipset play a role in connecting and coordinating communication
between different components. The quality and capabilities of the motherboard can affect the
overall system performance.
5. Graphics Processing Unit (GPU):
• In systems with dedicated GPUs, the graphics card's performance can impact tasks related
to graphics rendering, gaming, and certain computational workloads. Some applications leverage
GPU acceleration for improved performance.
6. Cooling and Thermal Management:
• Overheating can lead to performance throttling, where the CPU reduces its speed to
prevent damage. Efficient cooling solutions are essential for maintaining consistent performance
over extended periods.
7. Software Optimization:
• Well-optimized software can significantly impact system performance. Applications that
make efficient use of hardware resources can enhance overall responsiveness.
8. Operating System Efficiency:
• The efficiency of the operating system in managing resources, handling background
processes, and memory management can affect the overall system performance.
9. Peripheral Devices:
• The performance of input/output devices (e.g., keyboards, mice, and external drives) can
impact user experience, especially in tasks that involve a lot of data transfer.

POWER WALL
Pentium 4 made a dramatic jump in clock Rate and power but less in performance .Due to
thermal problem.

Core 2 has simpler pipeline with lower clock rates and multiple processors per chip.

IC ( Integrated circuit) are called CMOS ( Complementary metal oxide semi conductor)

For CMOS the Primary source of energy consumption is called dynamic energy.

Energy = capacity load * voltage 2