Jump to content

NAS Parallel Benchmarks

From Wikipedia, the free encyclopedia
This is an old revision of this page, as edited by Kxx (talk | contribs) at 00:03, 22 February 2009. The present address (URL) is a permanent link to this revision, which may differ significantly from the current revision.

The NAS Parallel Benchmarks are a set of benchmarks targetting performance evaluation of highly parallel supercomputers. They are developed and maintained by the NASA Advanced Supercomputing Division (formerly the NASA Numerical Aerodynamic Simulation Program) based at the NASA Ames Research Center.

Development

Motivation

The NAS Parallel Benchmarks were created in the early 1990s to address the lack of suitable benchmarks for highly parallel machines. Traditional benckmarks such as the Livermore loops, the LINPACK Benchmark and the original NAS Kernel Benchmark Program, being specialized for vector computers, suffered from inadequacies including parallelism-impeding tuning restrictions and insufficient problem sizes that rendered them inappropriate for highly parallel systems. Equally unsuitable were full-scale applications due to high porting cost.[1]

Guidelines

The specification of NAS Parallel Benchmarks recognizes the following as their development guidelines:

  • Use of new parallel-aware algorithmic and software methods
  • Genericness and architecture neutrality
  • Easy verifiability of correctness of results and performance figures
  • Capability of accomodating new systems with increased power
  • Ready distributability

References

  1. ^ D. Bailey, E. Barscz, J. Barton, D. Browning, R. Carter, L. Dagum, R. Fatoohi, S. Fineberg, P. Frederickson, T. Lasinski, R. Schreiber, H. Simon, V. Venkatakrishnan, S. Weeratunga, The NAS Parallel Benchmarks, NAS Technical Report RNR-94-007, NASA Ames Research Center, Moffett Field, CA, 1994.