Performance Analysis

What is the most appropriate approach when evaluating computer performance with benchmarks?

  1. AMeasure with programs close to the actual intended use and compare those results
  2. BDecide by the speed of a single small loop regardless of use
  3. CDecide by the manufacturer's published clock frequency alone
Show answer and explanation

Answer: A. Measure with programs close to the actual intended use and compare those results

Whether a computer is fast depends on what it is used for. Integer work, floating-point computation, memory-intensive processing and I/O-heavy workloads each favor different machines. So it is important to measure with programs as close as possible to actual use, or with standard benchmark suites collecting many representative programs, and to compare under matched conditions.

  • A. Correct. Which computer is faster depends on the use.
  • B. A small loop's results can differ greatly from real use.
  • C. Frequency alone does not determine performance.
Practice this topic Back to topic

More Performance Analysis questions

  1. Which describes a standard benchmark such as SPEC CPU?
  2. When combining ratios of speed against a reference machine across several benchmarks, why is the geometric mean often used?
  3. Two programs with equal instruction counts run back to back, one at 1000 million instructions per second and the other at 500 million. What is the overall average rate?
  4. Timing a small loop like the one above gave nearly zero time even for large N. What is the most likely cause?
  5. When timing a program, why are measurements often repeated with the first few runs discarded?