sharkdp/hyperfine

Reserve cpus for benchmarking

Open

#751 opened on Jun 19, 2024

View on GitHub
 (3 comments) (0 reactions) (0 assignees)Rust (28,110 stars) (479 forks)batch import
help wanted

Description

Hello, thank you for the great tool!

I was wondering whether it's better to combine this tool with something like cpuset, to reserve the cpus the benchmark and hyperfine run on.

Would this make a difference? I understand that hyperfine will catch outliers stemming from, e.g., competing programs, but would this be avoided in the first place by reserving the cpus?

Thank you for your time.

Contributor guide

Tech stack
rust
Domain
performance
Issue type
research
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
4
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
1-2 days
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
unclear
Prerequisites
None
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
30
Research direction
Explore the feasibility of integrating CPU pinning (e.g., using cpuset) into hyperfine to reduce benchmark variance. Review hyperfine's benchmarking loop in src/benchmark.rs and consider how to optionally set CPU affinity for the child process (e.g., via the nix crate or platform specific syscalls). Check if there are existing discussions or PRs about CPU isolation (none linked). Document the trade offs between using cpuset vs relying on hyperfine's statistical outlier detection.