SLIDE 1
- 4. Performance Analysis of
4.1 Performance Evaluation of Computer
User criteria:
- Small response times
Computing center criteria:
- High throughputs
4. Performance Analysis of Parallel Programs 4.1 Performance - - PowerPoint PPT Presentation
4. Performance Analysis of Parallel Programs 4.1 Performance - - PowerPoint PPT Presentation
4. Performance Analysis of Parallel Programs 4.1 Performance Evaluation of Computer User criteria: - Small response times Computing center criteria: - High throughputs 4.1.1 Evaluation of CPU Performance 4.1.1 Evaluation of CPU Performance
SLIDE 2
SLIDE 3
4.1.1 Evaluation of CPU Performance
SLIDE 4
4.1.1 Evaluation of CPU Performance
The response time of a program A can be split into:
SLIDE 5
User CPU time of A The response time of a program A can be split into:
4.1.1 Evaluation of CPU Performance
SLIDE 6
User CPU time of A System CPU time of A The response time of a program A can be split into:
4.1.1 Evaluation of CPU Performance
SLIDE 7
User CPU time of A System CPU time of A Waiting time of A The response time of a program A can be split into:
4.1.1 Evaluation of CPU Performance
SLIDE 8
User CPU time of A System CPU time of A Waiting time of A The response time of a program A can be split into:
4.1.1 Evaluation of CPU Performance
SLIDE 9
4.1.1 Evaluation of CPU Performance
User CPU time of A
SLIDE 10
User CPU time of A tcycle -> reciprocal to clock rate: T=1/f -> 2GHz = 1/(2*109)s = 0.5ns (cycle time) ncycle (A)-> total number of CPU cycles needed for all instructions of A
4.1.1 Evaluation of CPU Performance
SLIDE 11
CPI (Clock cycles Per Instruction)
4.1.1 Evaluation of CPU Performance
SLIDE 12
CPI (Clock cycles Per Instruction)
4.1.1 Evaluation of CPU Performance
SLIDE 13
CPI (Clock cycles Per Instruction) ninstr(A) -> total number of instructions executed for A
4.1.1 Evaluation of CPU Performance
SLIDE 14
ni(A) -> is the number of instructions of type Ii executed for the program A CPIi -> number of CPU cycles needed for instructions of type Ii CPI (Clock cycles Per Instruction)
4.1.1 Evaluation of CPU Performance
SLIDE 15
Example: We consider a processor with three instruction classes I1, I2, I3 containing instructions which require 1, 2, or 3 cycles for their execution. We assume that there are two difgerent possibilities for the translation
- f a
Programming language construct using difgerent instructions. CPI1 = 10/5 = 2 CPI2 = 9/6 = 1,5 CPI (Clock cycles Per Instruction)
4.1.1 Evaluation of CPU Performance
SLIDE 16
4.1.2 MIPS and MFLOPS
SLIDE 17
MIPS (Million Instructions Per Second)
4.1.2 MIPS and MFLOPS
SLIDE 18
Drawbacks/limitations:
- Only considers the number of instructions.
- MIPS rate does not necessarily correspond to the
execution time. MIPS (Million Instructions Per Second)
4.1.2 MIPS and MFLOPS
SLIDE 19
4.1.2 MIPS and MFLOPS
MFLOPS (Million Floating-point Operations Per Second)
SLIDE 20
4.1.2 MIPS and MFLOPS
MFLOPS (Million Floating-point Operations Per Second)
SLIDE 21
4.1.2 MIPS and MFLOPS
MFLOPS (Million Floating-point Operations Per Second) Drawbacks/limitations:
- Doesn’t difgerence between types of floating-points
- perations performed.
SLIDE 22
4.1.3 Performance of Processors with a Memory
SLIDE 23
4.1.3 Performance of Processors with a Memory
SLIDE 24
4.1.3 Performance of Processors with a Memory
nmm_cycles(A) -> number of additional machine cycles caused by memory accesses of A .