Processor Design

5Z032

The role of Performance

Henk Corporaal
Eindhoven University of Technology
2009

TU/e Processor Design 5Z032 1

 Performance
■ Measure, Report, and Summarize
■ Make intelligent choices
■ See through the marketing hype
■ Key to understanding underlying organizational motivation

Why is some hardware better than others for different programs?

What factors of system performance are hardware related?

(e.g., Do we need a new machine, or a new operating system?)

How does the Instruction Set Architecture (ISA) affect performance?

TU/e Processor Design 5Z032 2

 Which of these airplanes has the best
performance?
Airplane Passengers Range (mi) Speed (mph)

Boeing 737-100 101 630 598

Boeing 747 470 4150 610
BAC/Sud Concorde 132 4000 1350
Douglas DC-8-50 146 8720 544

 Which plane is the best?

• which metric do we use?
 How much faster is the Concorde compared to the 747?
 How much bigger is the 747 than the Douglas DC-8?
 Which plane has the best transporting capacity?

TU/e Processor Design 5Z032 3

 Computer
Performance:
TIME, TIME, TIME
■ Response Time, or Latency
— How long does it take for my job to run?
— How long does it take to execute a job?
— How long must I wait for the database query?
■ Throughput
— How many jobs can the machine run at once?
— What is the average execution rate?
— How much work is getting done?
■ Q1: If we upgrade a machine with a new processor what do we
increase?
■ Q2: If we add a new machine to the lab what do we increase?

TU/e Processor Design 5Z032 4

 Execution Time
■ Elapsed Time
◆ counts everything (disk and memory accesses, I/O , etc.)
◆ a useful number, but often not good for comparison purposes

■ CPU time
◆ doesn't count I/O or time spent running other programs
◆ can be broken up into system time, and user time

■ Our focus: user CPU time

◆ time spent executing the lines of code that are "in" our
program

TU/e Processor Design 5Z032 5

 Book's Definition of
Performance
For some program running on machine X,
■

PerformanceX = 1 / Execution timeX

"X is n times faster than Y"

■

PerformanceX / PerformanceY = n

Problem:
■ machine A runs a program in 20 seconds
machine B runs the same program in 25 seconds
◆
How much faster is A compared to B?
◆

TU/e Processor Design 5Z032 6

 Clock Cycles
■ Instead of reporting execution time in seconds, we often use
cycles seconds cycles seconds
= ×
program program cycle

■ Clock “ticks” indicate when to start activities (one abstraction):

cycle

■ cycle time = time between ticks = seconds per cycle

time
■ clock rate (frequency) = cycles per second (1 Hz. = 1
cycle/sec)

A 200 Mhz. clock has a

cycle time
1
× 10 9 = 5 nanoseconds
200 × 10 6

TU/e Processor Design 5Z032 7

 How to Improve
Performance
seconds cycles seconds
= ×
program program cycle

So, to improve performance (everything else being

equal) you can either

________ the # of required cycles for a program, or

________ the clock cycle time or, said another way,

________ the clock rate.

TU/e Processor Design 5Z032 8

 How to Improve
Performance
■ Put it another way:

CPU time = Ncycles * tclock = Ncycles / fclock

TU/e Processor Design 5Z032 9

 How many cycles are
required for a program?
■ Could assume that # of cycles = # of instructions

2nd instruction
3rd instruction
1st instruction

4th
5th
6th
...
time

This assumption is incorrect,

different instructions take different amounts of time on different machines.

Why?

TU/e Processor Design 5Z032 10

 Different numbers of cycles
for different instructions

time

 Multiplication takes more time than addition

 Floating point operations take longer than integer
ones
 Accessing memory takes more time than
accessing registers

 Important point: changing the cycle time often

changes the number of cycles required for various
instructions (more later)
TU/e Processor Design 5Z032 11
 Example
■ Our favorite program runs in 10 seconds on computer A, which has a
400 Mhz. clock. We are trying to help a computer designer build a
new machine B, that will run this program in 6 seconds.
The designer can use new (or perhaps more expensive) technology to
substantially increase the clock rate, but has informed us that this
increase will affect the rest of the CPU design, causing machine B to
require 1.2 times as many clock cycles as machine A for the same
program.

■ What clock rate should we tell the designer to target?"

(Don't Panic, you can easily work this out from basic principles !!)

TU/e Processor Design 5Z032 12

 Now that we
understand cycles
■ A given program will require
◆ some number of instructions (machine instructions)
◆ some number of cycles
◆ some number of seconds
■ We have a vocabulary that relates these quantities:
◆ cycle time (seconds per cycle)
◆ clock rate (cycles per second)
◆ CPI (cycles per instruction)
a floating point intensive application might have a higher CPI
◆ MIPS (millions of instructions per second)
this would be higher for a program using simple instructions

TU/e Processor Design 5Z032 13

 Performance
■ Performance is determined by execution time
■ Do any of the other variables equal performance?

◆ # of cycles to execute program?

◆ # of instructions in program?
◆ # of cycles per second?
◆ average # of cycles per instruction?
◆ average # of instructions per second?

■ Common pitfall: thinking one of the variables is

indicative of performance when it really isn’t.

TU/e Processor Design 5Z032 14

 Performance
■ Famous equation:

texec = N instructions • CPI • tclock

note that : N cycles = N instructions • CPI

TU/e Processor Design 5Z032 15

 CPI Example
■ Suppose we have two implementations of the same
instruction set architecture (ISA).

For some program,

Machine A has a clock cycle time of 10 ns. and a CPI of

2.0
Machine B has a clock cycle time of 20 ns. and a CPI of
1.2

What machine is faster for this program, and by how

much?

■ If two machines have the same ISA which of our quantities (e.g.,
clock rate, CPI, execution time, # of instructions, MIPS) will
always be identical?

TU/e Processor Design 5Z032 16

 # of Instructions
Example
■ A compiler designer is trying to decide between two code
sequences for a particular machine. Based on the
hardware implementation, there are three different classes
of instructions: Class A, Class B, and Class C, and they
require one, two, and three cycles (respectively).

The first code sequence has 5 instructions:

2 of A, 1 of B, and 2 of C
The second sequence has 6 instructions:
4 of A, 1 of B, and 1 of C.

Which sequence will be faster? How much?

What is the CPI for each sequence?

TU/e Processor Design 5Z032 17

 MIPS example
Two different compilers are being tested for a 100 MHz. machine
with three different classes of instructions: Class A, Class B, and
Class C, which require one, two, and three cycles (respectively).
Both compilers are used to produce code for a large piece of
software.
The first compiler's code uses 5 million Class A instructions, 1
million Class B instructions, and 1 million Class C instructions.
The second compiler's code uses 10 million Class A instructions,
1 million Class B instructions, and 1 million Class C instructions.
■ Which sequence will be faster according to MIPS?
■ Which sequence will be faster according to execution time?

TU/e Processor Design 5Z032 18

 Benchmarks
■ Performance best determined by running a real application
◆ Use programs typical of expected workload
◆ Or, typical of expected class of applications
e.g., compilers/editors, scientific applications, graphics, etc.
■ Small benchmarks
◆ nice for architects and designers
◆ easy to standardize
◆ can be abused
■ SPEC (System Performance Evaluation Cooperative)
◆ companies have agreed on a set of real program and inputs
◆ can still be abused (Intel’s “other” bug)
◆ valuable indicator of performance (and compiler technology)

TU/e Processor Design 5Z032 19

 SPEC ‘95
Be n ch m a rk De scriptio n
go A rtific ial intelligenc e; play s the gam e of G o
m 88k s im M otorola 88k c hip s im ulator; runs tes t program
gc c The G nu C c om piler generating S P A RC c ode
C c om pres s
li
Com pres s es and dec om pres s es file in m em ory
Lis p interpreter
ijpeg G raphic c om pres s ion and dec om pres s ion
perl M anipulates s trings and prim e num bers in the s pec ial-purpos e progr. lan
vortex A databas e program
tom c atv A m es h generation program
s wim S hallow water m odel with 513 x 513 grid
s u2c or quantum phy s ic s ; M onte Carlo s im ulation
hy dro2d A s trophy s ic s ; Hy drody nam ic Naiver S tok es equations
Fortran m grid M ultigrid s olver in 3-D potential field
applu P arabolic /elliptic partial differential equations
trub3d S im ulates is otropic , hom ogeneous turbulenc e in a c ube
aps i S olves problem s regarding tem perature, wind veloc ity , and dis tribution o
fpppp Q uantum c hem is try
wave5 P las m a phy s ic s ; elec trom agnetic partic le s im ulation

TU/e Processor Design 5Z032 20

 SPEC ‘95 Integer
Does doubling the clock rate double the performance?
Can a machine with a slower clock rate have better performance?

10

6
SPECint

0
50 100 150 200 250

Clock rate (MHz) Pentium

Pentium Pro

TU/e Processor Design 5Z032 21

 SPEC`95: FP
10

6
SPECfp

0
50 100 150 200 250
Clock rate (MHz)
Pentium

Pentium Pro

TU/e Processor Design 5Z032 22

 Amdahl's Law
Execution Time After Improvement =

Execution Time Unaffected +( Execution Time Affected / Amount of

Improvement )

■ Example:

"Suppose a program runs in 100 seconds on a machine, with

the multiply operation responsible for 80 seconds of this time.
How much do we have to improve the speed of multiplication if
we want the program to run 4 times faster?"

How about making it 5 times faster?

■ Principle: Make the common case fast

TU/e Processor Design 5Z032 23

 Examples
■ Suppose we enhance a machine making all floating-point instructions run
five times faster. If the execution time of some benchmark before the
floating-point enhancement is 10 seconds, what will the speedup be if half
of the 10 seconds is spent executing floating-point instructions?

■ We are looking for a benchmark to show off the new floating-point unit
described above, and want the overall benchmark to show a speedup of 3.
One benchmark we are considering runs for 100 seconds with the old
floating-point hardware. How much of the execution time would floating-
point instructions have to account for in this program in order to yield our
desired speedup on this benchmark?

TU/e Processor Design 5Z032 24

 Remember
■ Performance is specific to a particular program (and its data
set)
◆ Total execution time is a consistent summary of performance

■ For a given architecture performance increases come from:

◆ increases in clock rate (without adverse CPI affects)
◆ improvements in processor organization that lower CPI
◆ compiler enhancements that lower CPI and/or instruction count

■ Pitfall: expecting improvement in one aspect of a machine’s

performance to affect the total performance
■ You should not always believe everything you read! Read
carefully!
(see newspaper articles, e.g., Exercise 2.37)

TU/e Processor Design 5Z032 25

 Review: Aspects of CPU
Performance
CPU
CPUtime
time == Seconds
Seconds ==Instructions
Instructions xx Cycles
Cycles xx Seconds
Seconds
Program
Program Program
Program Instruction
Instruction Cycle
Cycle

instr count CPI clock rate

Program X
Compiler X X
Instr. Set X X
Organization X X
Technology X

TU/e Processor Design 5Z032 26

 Exercises
From Chapter two (2nd ed.):

■ 2.1, 2.2, 2.5, 2.8

■ 2.10 - 2.13

■ 2.18 - 2.24

TU/e Processor Design 5Z032 27