This document discusses hardware and software parallelism in computer systems. It defines hardware parallelism as parallelism enabled by the machine architecture through multiple processors or functional units. Software parallelism refers to parallelism exposed in a program's control and data dependencies. Modern computer architectures require support for both types of parallelism to perform multiple tasks simultaneously. However, there is often a mismatch between the hardware and software parallelism available. For example, a dual-processor system may be able to execute 12 instructions in 6 cycles, but the program's inherent parallelism may only allow completing the instructions in 7 cycles. Achieving optimal parallelism requires coordination between hardware design and software programming.

1. 1
Hardware and Software Parallelism
Dept. of Computer Science & Engineering
2013-2014
Presented by
Prashant Dahake
Mtech 1st
sem (CSE)
Sub:- High Performance Computer
Architecture
1
A
Presentation on
G.H. Raisoni College of Engineering Nagpur
1

2. What is Parallelism ?
 Performing more than one task at a time.
 It speed up the process.
 More work in less time .
 It reduces the cost of work.

3.  Modern computer architecture implementation
requires special hardware and software support for
parallelism.
Types of parallelism
 Hardware parallelism
 Software parallelism

4. Hardware Parallelism:
 This refers to the type of parallelism defined by the machine
architecture and hardware multiplicity.
 Hardware parallelism is a function of cost and performance tradeoffs. It
displays the resource utilization patterns of simultaneously executable
operations. It can also indicate the peak performance of the processors.
 One way to characterize the parallelism in a processor is by the number
of instruction issues per machine cycle.
 In a modern processor, two or more instructions can be issued per
machine cycle. e .g i960CA was a 3-issue processor.

5. Software Parallelism:
 It is defined by the control and data dependence of programs.
 The degree of parallelism is revealed in the program profile or in the
program flow graph.
 Software parallelism is a function of algorithm, programming style, and
compiler optimization.
 The program flow graph displays the patterns of simultaneously
executable operations.
 Parallelism in a program varies during the execution period. It limits the
continuous performance of the processor.

6. Mismatch between H/W and S/W parallelism
Example:
A= L1*L2 + L3*L4
B= L1*L2 - L3*L4
Software Parallelism
There are 8 instructions;
FOUR Load instructions (L1, L2, L3 & L4).
TWO Multiply instructions (X1 & X2).
ONE Add instruction (+)
ONE Subtract instruction (-)
The parallelism varies from 4 to
three cycles.
Average s/w parallelism = 8/3 =2.67

7. Hardware Parallelism:
Parallel Execution:
•
•
Using TWO-issue processor:
The processor can execute
one memory access (Load
or Store) and one arithmetic
operation (multiply, add,
subtract) simultaneously.
The program must execute
in cycles.
• 7
• The h/w parallelism average
is 8/7=1.14.
This demonstrate the
mismatch between h/w and
s/w parallelism.
•
G

8. Example:
 A h/w platform of a Dual-Processor system,
single issue processors are used to execute the
same program.
 Six processor cycles are needed to execute the
12 instructions by two processors.
 S1 & S2 are two inserted store operations.
 L5 and L6 are two inserted load operation.
 The added instructions are needed for inter-
processor communication through the shared
memory

9. Conclusion
To achieve parallelism joint efforts between
hardware designer and software programmer
are needed which further upgrade computer
performance.

10. Thank you