A5 Solution

ICS 233 – Computer Architecture
& Assembly Language

Assignment 5: Performance and Metrics
Solution
1. (4 pts) We wish to compare the performance of two different computers: M1 and M2. The
following measurements have been made on these computers:
Program Time on M1 Time on M2

1 2.0 seconds 1.5 seconds
2 5.0 seconds 10.0 seconds
Program Instructions executed on M1 Instructions executed on M2

1 5 × 109 6 × 109
a) (1 pt) Which computer is faster for each program, and how many times as fast is it?
b) (1 pt) Find the instruction execution rate (instructions per second) for each computer
when running program 1.
c) (1 pt) The clock rates for M1 and M2 are 3 GHz and 5 GHz respectively. Find the CPI
for program 1 on both machines.
d) (1 pt) Suppose that program 1 must be executed 1600 times each hour. Any remaining
time should used to run program 2. Which computer is faster for this workload?
Performance is measured here by the throughput of program 2.
Solution:
a) For program 1, M2 is 2.0/1.5 = 1.33 times as fast as M1.
For program 2, M1 is 10.0/5.0 = 2 times as fast as M2.
b) For program 1:
Execution rate on M1 = 5 × 109 / 2.0 = 2.5 × 109 IPS (Instructions Per Second).
Execution rate on M2 = 6 × 109 / 1.5 = 4 × 109 IPS.
c) CPI = Execution time × Clock rate / Instruction Count

For program 1:
CPI on M1 = 2.0 × 3 × 109 / (5 × 109) = 1.2 cycles per instruction
CPI on M2 = 1.5 × 5 × 109 / (6 × 109) = 1.25 cycles per instruction
d) Running program 1 1600 times each hour:

On M1, time required for program 1 = 1600 × 2.0 = 3200 seconds
On M2, time required for program 1 = 1600 × 1.5 = 2400 seconds
Time left to run program 2 each hour:

On M1, time left for program 2 = 3600 – 3200 = 400 seconds
On M2, time left for program 2 = 3600 – 2400 = 1200 seconds
Prepared by Dr. Muhamed Mudawar Page 1 of 4

In that time left, program 2 can run:
On M1, program 2 can run 400/5 = 80 times
On M2, program 2 can run 1200/10 = 120 times
Thus M2 performs better on this workload than M1.
2. (2 pts) Suppose you wish to run a program P with 7.5 × 109 instructions on a 5 GHz
machine with a CPI of 1.2.
a) (1 pt) What is the CPU execution time?
b) (1 pt) When you run program P, it takes 3 seconds of wall time to complete. What is
the percentage of the CPU time program P received?
Solution:
a) CPU execution time = Instruction Count × CPI / Clock rate

CPU execution time = 7.5 × 109 × 1.2 / (5 × 109) = 1.8 seconds
b) % of CPU time = 1.8 / 3 = 0.6 or 60% of the total wall time
3. (4 pts) Consider two different implementations, M1 and M2, of the same instruction set.
There are five classes of instructions (A, B, C, D and E) in the instruction set. M1 has a
clock rate of 4 GHz and M2 has a clock rate of 6 GHz.
Class CPI on M1 CPI on M2

A 1 2
B 2 2
C 3 2
D 4 4
E 3 4
a) (2 pts) Assume that peak performance is defined as the fastest rate that a computer
can execute any instruction sequence. What are the peak performances of M1 and M2
expressed in instructions per second?
b) (2 pts) If the number of instructions executed in a certain program is divided equally
among the classes of instructions, except that for class A, which occurs twice as often
as each of the others, how much faster is M2 than M1?
Solution:
a) For peak performance, the machine will be executing the fastest set of
instructions. M1 will be executing instructions of class A only, and M2 will be
executing instructions that belong to class A, B, or C.
Peak performance of M1 = 4 × 109 / 1 = 4 × 109 IPS = 4000 MIPS

Peak performance of M2 = 6 × 109 / 2 = 3 × 109 IPS = 3000 MIPS

b) Average CPI on M1 = (2 × 1 + 2 + 3 + 4 + 3) / 6 = 14 / 6 = 2.33
Average CPI on M2 = (2 × 2 + 2 + 2 + 4 + 4) / 6 = 16 / 6 = 2.67
Since the instruction count is the same:
Performance of M2 Clock Rate of M2 CPI on M1

= ×
Performance of M1 Clock Rate of M1 CPI on M2
M2 is faster than M1 by a factor of (6 GHz / 4 GHz) × (2.33 / 2.67) = 1.31
4. (5 pts) Consider two different implementations, M1 and M2, of the same instruction set.
There are three classes of instructions (A, B, and C) in the instruction set. M1 has a clock
rate of 6 GHz and M2 has a clock rate of 3 GHz. The CPI for each instruction class on
M1 and M2 is given in the following table:
Class CPI on M1 CPI on M2 C1 Usage C2 Usage C3 Usage

A 2 1 40% 40% 60%
B 3 2 40% 20% 15%
C 5 2 20% 40% 25%
The above table also contains a summary of the usage of instruction classes generated by
three different compilers: C1, C2, and C3. Assume that each compiler generates the same
number of instructions for a given program.
a) (1 pt) Using C1 compiler on both M1 and M2, how much faster is M1 than M2?
b) (1 pt) Using C2 compiler on both M1 and M2, how much faster is M2 than M1?
c) (1 pt) If you purchase M1, which compiler would you use?
d) (1 pt) If you purchase M2, which compiler would you use?
e) (1 pt) Which computer and compiler combination give the best performance?
Solution:
a) Using C1 compiler:
Average CPI on M1 = 0.4 × 2 + 0.4 × 3 + 0.2 × 5 = 3.0
Average CPI on M2 = 0.4 × 1 + 0.4 × 2 + 0.2 × 2 = 1.6
b) Using C2 compiler:
Average CPI on M1 = 0.4 × 2 + 0.2 × 3 + 0.4 × 5 = 3.4
Average CPI on M2 = 0.4 × 1 + 0.2 × 2 + 0.4 × 2 = 1.6
c) Using C3 compiler on M1:

Average CPI on M1 = 0.6 × 2 + 0.15 × 3 + 0.25 × 5 = 2.9
We should use compiler C3 because it gives the lowest CPI on M1.

d) Using C3 compiler on M2:
Average CPI on M2 = 0.6 × 1 + 0.15 × 2 + 0.25 × 2 = 1.4
We should use compiler C3 because it gives the lowest CPI on M2.
e) Compiler C3 compiler gives the best average CPI on both M1 and M2.
Performance is proportional to Clock rate / CPI, because I-count is the same.
Performance of M1 relative to M2 = (6 GHz / 3 GHz) × (1.4 / 2.9) = 2.8/2.9
Therefore, M2 gives the best performance using Compiler C3.
5. (2 pts) A benchmark program runs for 100 seconds. We want to improve the speedup of
the benchmark by a factor of 3. We enhance the floating-point hardware to make floating-
point instructions run 5 times faster. How much of the initial execution time would
floating-point instructions have to account for to show an overall speedup of 3 on this
benchmark?
Let f be the fraction of time spent in floating point instructions. Then, after the
enhancement, the time taken by the machine will be:
100 × (f / 5 + (1 – f)) = 100 / 3

(f / 5 + (1 – f)) = 1 / 3
1 – 1/3 = f – f / 5
4 f /5 = 2 / 3
f = (2 / 3) (5 / 4) = 5 / 6
Therefore, 5/6 of the initial 100 seconds, or 83.33 seconds, must be spent
executing floating-point instructions.
6. (3 pts) Consider the following fragment of MIPS code. Assume that a and b are arrays of
words and the base address of a is in $a0 and the base address of b is in $a1. How many
instructions are executed during the running of this code? If ALU instructions (addu and
addiu) take 1 cycle to execute, load/store (lw and sw) take 5 cycles to execute, and the
branch (bne) instruction takes 3 cycles to execute, how many cycles are needed to
execute the following code (all iterations). What is the average CPI?
addu $t0, $zero, $zero # i = 0

addu $t1, $a0, $zero # $t1 = address of a[i]
addu $t2, $a1, $zero # $t2 = address of b[i]
addiu $t3, $zero, 101 # $t3 = 101 (max i)
loop: lw $t4, 0($t2) # $t4 = b[i]
addu $t5, $t4, $s0 # $t5 = b[i] + c
sw $t5, 0($t1) # a[i] = b[i] + c
addiu $t0, $t0, 1 # i++
addiu $t1, $t1, 4 # address of next a[i]
addiu $t2, $t2, 4 # address of next b[i]
bne $t0, $t3, loop # loop if (i != 101)
Total number of instructions executed = 4 + 101 × 7 = 711.

ALU = 4 + 101 × 4 = 408, lw and sw = 101 × 2 = 202, and bne = 101 instructions.
Total number of cycles = 408 × 1 + 202 × 5 + 101 × 3 = 1721.
Average CPI = 1721 / 711 = 2.42

A5 Solution

Uploaded by

Copyright:

Available Formats

A5 Solution

Uploaded by

Document Information

Original Description:

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

A5 Solution

Uploaded by

Copyright:

Available Formats

ICS 233 – Computer Architecture

& Assembly Language

Program Time on M1 Time on M2

Program Instructions executed on M1 Instructions executed on M2

c) CPI = Execution time × Clock rate / Instruction Count

d) Running program 1 1600 times each hour:

Time left to run program 2 each hour:

Prepared by Dr. Muhamed Mudawar Page 1 of 4

Thus M2 performs better on this workload than M1.

a) CPU execution time = Instruction Count × CPI / Clock rate

b) % of CPU time = 1.8 / 3 = 0.6 or 60% of the total wall time

Class CPI on M1 CPI on M2

Peak performance of M1 = 4 × 109 / 1 = 4 × 109 IPS = 4000 MIPS

Prepared by Dr. Muhamed Mudawar Page 2 of 4

Since the instruction count is the same:

Performance of M2 Clock Rate of M2 CPI on M1

M2 is faster than M1 by a factor of (6 GHz / 4 GHz) × (2.33 / 2.67) = 1.31

Class CPI on M1 CPI on M2 C1 Usage C2 Usage C3 Usage

M1 is faster than M2 by a factor of (6 GHz / 3 GHz) × (1.6 / 3.0) = 1.0667

M2 is faster than M1 by a factor of (3 GHz / 6 GHz) × (3.4 / 1.6) = 1.0625

c) Using C3 compiler on M1:

Prepared by Dr. Muhamed Mudawar Page 3 of 4

100 × (f / 5 + (1 – f)) = 100 / 3

addu $t0, $zero, $zero # i = 0

Total number of instructions executed = 4 + 101 × 7 = 711.

Prepared by Dr. Muhamed Mudawar Page 4 of 4

You might also like