NVIDIA/nvbench
1
0
Fork 0
mirror of https://github.com/NVIDIA/nvbench.git synced 2026-04-20 14:58:54 +00:00
Code Issues Packages Projects Releases Wiki Activity

Update best_practices.md

Browse Source
This commit is contained in:
Yunsong Wang
2025-09-19 12:58:37 -07:00
committed by GitHub
parent df27dcb5df
commit 54cbcd0c42
1 changed files with 1 additions and 2 deletions
Download Patch File Download Diff File Expand all files Collapse all files

3
docs/best_practices.md
View File

@@ -6,8 +6,7 @@ This document is **not intended to replace** the detailed benchmark documentatio
## NVBench
* Purpose-built for CUDA GPU workloads.
* Provides GPU-aware features: warmup runs, synchronization, throughput/latency metrics, and parameter sweeps, etc.
* Provides GPU-aware features: runtime customization, throughput calculations, and parameter sweeps, etc.
* Produces machine-readable output (JSON, CSV) suitable for regression tracking and CI pipelines.
* The natural choice for benchmarking GPU-accelerated code.
* Also supports benchmarking normal CPU implementations.