English

Robust benchmarking in noisy environments

Performance 2016-08-16 v1

Abstract

We propose a benchmarking strategy that is robust in the presence of timer error, OS jitter and other environmental fluctuations, and is insensitive to the highly nonideal statistics produced by timing measurements. We construct a model that explains how these strongly nonideal statistics can arise from environmental fluctuations, and also justifies our proposed strategy. We implement this strategy in the BenchmarkTools Julia package, where it is used in production continuous integration (CI) pipelines for developing the Julia language and its ecosystem.

Keywords

Cite

@article{arxiv.1608.04295,
  title  = {Robust benchmarking in noisy environments},
  author = {Jiahao Chen and Jarrett Revels},
  journal= {arXiv preprint arXiv:1608.04295},
  year   = {2016}
}

Comments

7 pages, 5 figures, Proceedings of the 20th Annual IEEE High Performance Extreme Computing Conference, 2016

R2 v1 2026-06-22T15:20:01.279Z