Benchmark Skepticism

The cluster focuses on debates about the reliability, reproducibility, manipulation, and real-world relevance of performance benchmarks in software and hardware contexts.

➡️ Stable 0.8x Hardware
4,717
Comments
20
Years Active
5
Top Authors
#4400
Topic ID

Activity Over Time

2007
2
2008
21
2009
65
2010
114
2011
104
2012
164
2013
227
2014
216
2015
207
2016
188
2017
282
2018
306
2019
255
2020
338
2021
434
2022
366
2023
396
2024
422
2025
578
2026
32

Keywords

SLO AAA ARC TBH benchmarks benchmark benchmarking performance results general real performs conclusions drawing

Sample Comments

ladyanita22 • Dec 11, 2023 • View on HN

Well, the benchmarks don't tell that.

TheRealKing • Oct 18, 2018 • View on HN

Wouldn't it depend on how those benchmarks were written?

flir • Mar 25, 2025 • View on HN

Man, I hope those benchmarks actually measure something.

paulryanrogers • Aug 22, 2024 • View on HN

Does benchmark-oriented means it's unlikely to help (or could even hurt) real world applications?

monster_truck • Oct 16, 2025 • View on HN

Spend some time looking at the real benchmarks before writing nonsense

voz_ • Jul 30, 2022 • View on HN

This doesn't mean much without benchmarks

Doesn't that just kinda lend further support to the benchmarks being a bit silly?

ig-88ms • Jun 18, 2022 • View on HN

Don't trust benchmarks. They're all manipulated by the hardware companies

pixelpoet • Mar 12, 2025 • View on HN

Not a single benchmark on the page...Just sloppy TBH.

patronagezero • Feb 15, 2015 • View on HN

Benchmarks are snake oil when not being given the ability to reproduce them.