Code performance measurement as an educational tool

  • Authors:
  • Keith Schoby;Nelson Passos;Stewart Carpenter

  • Affiliations:
  • Midwestern State University, Wichita Falls, TX;Midwestern State University, Wichita Falls, TX;Midwestern State University, Wichita Falls, TX

  • Venue:
  • Journal of Computing Sciences in Colleges
  • Year:
  • 2007

Quantified Score

Hi-index 0.00

Visualization

Abstract

Physical restraints on current processor technologies are limiting the potential for continued processor speed improvements. Next-generation processor technologies might be decades off or might never come to fruition. As the increase in processor speed slows from the dramatic pace of Moore's Law, the efficiency of the running software is again becoming important. Is code performance measurement a valid topic in a computer science department's curriculum? Can students measure the performance of the programs they write? This study attempts to answer these questions by evaluating the ease with which a performance measurement tool can be applied to a simple programming assignment, and whether it can discern performance differences in small student programs as well as differences in processors.