TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Compilers Love Messing With Benchmarks

27 pointsby helperabout 11 years ago

2 comments

habosaabout 11 years ago
Measurement bias can have a huge effect, even when you think you are doing &#x27;good&#x27; benchmarking.<p>See the paper &#x27;How To Produce Wrong Data Without Doing Anything Obviously Wrong!&#x27; [1]. They just change the size of environment variables and the link order of files, and get wildly different performance results. Most people just assume if they close all other running programs they&#x27;ll have a fair benchmark, but there is so much more to consider.<p>[1] - <a href="http://www-plan.cs.colorado.edu/diwan/asplos09.pdf" rel="nofollow">http:&#x2F;&#x2F;www-plan.cs.colorado.edu&#x2F;diwan&#x2F;asplos09.pdf</a>
pdknskabout 11 years ago
The only conclusion here is not to use old and&#x2F;or synthetic benchmarks.
评论 #7695347 未加载
评论 #7695713 未加载