Comparch 06 Advanced Concepts
Comparch 06 Advanced Concepts
ADVANCED
CONCEPTS
1
Computer
Architecture and
Organization
Computer Architecture and Organization
LEARNING OBJECTIVES
Discuss the principle behind instruction pipelining and how
it works in practice.
hazards.
Computing
Computer Architecture and Organization 2
INSTRUCTION
PIPELINING
Is similar to an assembly line in a manufacturing plant.
products at various stages can be worked on simultaneously.
new inputs are accepted at one end before previously
accepted inputs appear as outputs at the other end.
Instruction pre-fetch / fetch overlap :- to fetch the next
instruction in parallel with the execution of the current
one.
Pipelining requires registers to store data between
stages. (involves instructions buffering)
This process will speed up instruction execution. (if equal
duration stages)
ADD R1, R2
SUB R3, R1
There are three types of data hazards: RAW, WAR and WAW
Read after write (RAW), or true dependency: An instruction
modifies a register or memory location and a succeeding
instruction reads the data in that memory or register location.
A hazard occurs if the read takes place before the write operation is
complete.
This is the most common type of hazard and the kind that we use
forwarding to overcome
Clock cycle
Clock cycle
1 2 3 4 5 6 7 8 9
1 2 3 4 5 6 7 8 9
I1 FI DI FO EI WO
I1 FI DI FO EI WO
I2 FI DI FO EI WO
I2 FI DI Idle FO EI WO
I3 FI DI FO EI WO
I3 FI DI Idle FO EI WO
I4 FI DI FO EI WO
I4 FI DI FO EI WO
DEALING WITH
1 ADD R1, R2
2 BRZ 7
3 SUB R4, R1
BRANCHES
Multiple streams
⁞
7
⁞
MOV R4,
10010
1 2 3 4 5 6 7 8 9 10 11 12 1 2 3 4 5 6 7 8 9
I1 FI DI FO EI WO I1 FI DI FO EI WO
I2 FI DI FO EI WO I2 FI DI FO EI WO
I3 FI DI FO EI WO I3 FI DI FO EI WO
I4 FI DI FO EI WO I7 FI DI FO EI WO
I5 FI DI FO EI WO I8 FI DI FO EI WO
I6 FI DI FO EI WO
I7 FI DI FO EI WO Pipeline stream 2
I8 FI DI FO EI WO
(by assuming the branch is taken)
17
Pipeline stream 1 (by assuming the branch not taken)
DEALING WITH
BRANCHES
Pre-fetch branch target
fetched.
Used in IBM 360/91
local memory.
Communication between processors occurs through message passing,
high-performance computing.
Computer Architecture and Organization 29
TYPES OF PARALLEL
PROCESSOR SYSTEMS
ARCHITECTURE
Shared Memory MIMD architectures can be further classified into
banks, and each processor has faster access to its local memory
compared to remote memory.
This results in non-uniform memory access times.
software.
2. Memory Management:
The OS is responsible for managing the computer's memory. It
4. Device Management:
The OS manages the computer's input/output devices, such as
6. Security:
The OS ensures the security and protection of the system and