+1 vote
in Computer Architecture by
Processor performance ‘P’ could be stated to linearly relate to the time ‘T’it takes to run an application program. State the relations between the performance of a processor and Instructions/Program, Cycles per Instructions and Time per cycle. Discuss the limitation of each of these individual components or parts of this equation to the overall performance of a processor as seen from computer architecture design.

1 Answer

+1 vote
by

performance = time/program

=instructions/program × cycles/instruction × time/cycle

(code size)                           (CPI)                    (cycle time)

CISC performs multiple instructions per cycle and RISC only one.

Using CISC and RISC as basis for comparison;

  • considering only instructions per program, keeping the rest constant, CISC would perform better than RISC
  • considering cycles per instruction and keeping the rest constant, RISC would perform better than CISC
  • increment in frequency increases performance although a great speedup will nullify that effect (frequency is inverse of cycle time)

Related questions

Welcome to CPEN Talk
Solution-oriented students of computer engineering on one platform to get you that

ONE SOLUTION

Chuck Norris insists on strongly-typed programming languages.
...