Lecture - 05
Performance of a Computer
Response Time
- The time between the start and completion of a
task.
It is also known as execution time.
- This includes disk accesses, memory accesses,
I/O activities, operating system overhead, CPU
execution time, etc.
Throughput
- The total amount of job done in a fixed amount
of time.
- Decreasing response time almost always
Relation between Performance and
Execution Time
PerformanceX =1/ Execution timeX
For two computers X and Y, if the performance
of X is greater than the performance of Y, then
PerformanceX > PerformanceY
= 1/ Execution timeX > 1/ Execution timeY
= Execution timeY > Execution timeX
If X is n times faster than Y,
then PerformanceX /
PerformanceY = n
Relative Performance
If computer A runs a program in 10 seconds and
computer B runs the same program in 15 seconds,
how much faster is A than B?
PerformanceA / PerformanceB
= Execution timeB / Execution timeA
= 15/10
= 1.5
A is 1.5 times faster than B.
Measuring Performance
Time is the measure of computer performance.
Computers are often shared. In such cases, the
system may try to optimize throughput rather
than minimizing the elapsed time for one
program.
CPU execution time or CPU time is the time the
CPU spends computing a task not including the
time spent waiting for I/O or running other
programs.
Measuring Performance
CPU time- 1. User CPU time
2. System CPU time
CPU time spent in the program is called User
CPU time, and
The CPU time spent in the operating system
performing tasks requested by the program is
called System CPU time.
Measuring Performance
Suppose, after executing a program we get the
following measurements:
User CPU time – 90.7 seconds
System CPU time – 12.9
seconds Elapsed time – 159
seconds
CPU Utilization is
(90.7+12.9)/159 = 65%
• about 65.2% of the elapsed time, the
CPU was actively processing the
program
The term system performance is used to refer to
elapsed time on an unloaded system while CPU
Clock Cycles
Almost all computers are constructed using a
clock that runs at a constant rate and determines
when events take place in the hardware.
Clock period is the time for a complete clock
cycle.
Clock rate is the inverse of clock period.
CPU Time
CPU time for a program can be expressed in
two ways: CPU time = CPU clock cycles for a program *
clock cycle time or
CPU time = CPU clock cycles for a program / Clock rate
If we know the number of clock cycles and the
number of instructions execute (Instruction count,
IC), we can calculate the average number of clock
cycles per instruction (CPI).
CPI = CPU clock cycles for a program / Instruction count
Instructions per clock (IPC) is the inverse of CPI.
CPU Time
CPU time = Instruction count * Cycles per instruction * Clock
cycle time
or
CPU time = Instruction count * Cycles per instruction / Clock rate
T = N * CPI / f
It is possible to compute the CPU clock cycles by
looking at the different types of instructions and
using their individual clock cycle counts.
CPU clock cycles = (CPIi X Ci)
where Ci is the count of the number of instructions
of class i executed, CPIi is the average number of
cycles per instruction for that instruction class,
Comparing Code Segments
CPI for this instruction class Code Instruction count for instruction class
A B C Sequence A B C
CPI 1 2 3
1 2 1 2
2 4 1 1
a) Which code sequence executes the most instructions?
b) Which will be faster?
c) What is the CPI for each sequence?
Sequence 1 executes 5 instructions and sequence 2
executes 6 instructions.
CPU clock cycles1 = (2X1)+(1X2)+(2X3)
= 10 cycles CPU clock cycles2 =
(4X1)+(1X2)+(1X3) = 9 cycles So code
sequence 2 is faster.
Since sequence 2 takes fewer overall clock cycles but
has more instruction, it must have a lower CPI.
CPI1 = 10/5 =
Practice Problem
Our favorite program runs in 10 seconds on computer A, which has a 2 GHz
clock. We are trying to help a computer designer build a computer, B, which will
run this program in 6 seconds. The designer has determined that a substantial
increase in the clock rate is possible, but this increase will affect the rest of the
CPU design, causing computer B to require 1.2 times as many clock cycles as
computer A for this program. What clock rate should we tell the designer to
target?
CPU Time
CPU performance depends on three
characteristics:
- Clock cycle time (Clock rate) : Hardware
technology
- Clock cycles per instruction: Architecture
- Instruction count (N) : software technology
Components affect the CPU
Performance
Hardware or software Affects what?
component
Algorithm Instruction count
Programming language Instruction
count CPI
Compiler Instruction
count CPI
Instruction set Instruction
architecture count Clock
rate
CPI
Speedup Techniques
Feature Objective
Cache To provide the CPU with faster access to
Memor instruction and data.
y
Pipelined To increase performance by allowing the
Processin processing of several instructions to be partially
g overlapped.
Superscal To increase performance by allowing
ar several instructions to be processed in
Processing parallel (full overlapping).
Sequential and Pipelined Processing
