NVIDIA / NVIDIA/cuvs

[FEA] Add CPU and GPU telemetry to cuVS Bench

Open
#1,624 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

feature request
Dominant language
Cuda
Stars
854
Forks
236
Avg merge
3d 3h
Merged PRs (30d)
62

Description

Request to add the following metric per grid point in cuVS Bench-

  1. Max RAM / VRAM (in GB)
  2. Avg CPU / GPU %

The above should be done seperately for the index build and search phase... so a total of 16 additional columns: max_ram_build_gb, max_ram_search_gb, max_vram_build_gb, max_vram_build_gb
avg_cpu_perc_build, avg_gpu_perc_build, avg_cpu_perc_search, avg_gpu_perc_search.

This would help with-

  1. Tradeoffs between search / build perf and memory.
  2. Biggest dataset that can be used to build cagra index for a specific GPU SKU.

Contributor guide

Open the contributing guide

First steps

  1. Read the whole issue, then the project's contributing guide.
  2. Comment on the issue to say you are picking it up — it saves two people doing the same work.
  3. Fork the repository and make your change on a branch.
  4. Open a pull request that references the issue number.

Research direction

Start by locating the cuVS Bench implementation and the code paths for index build and search measurements. Determine how per-grid-point results are collected and exported, then define the complete set of RAM, VRAM, CPU, and GPU columns for both phases. Done means the telemetry appears separately for build and search in benchmark output and is validated on representative runs.

Written by the indexing model from the issue text.

Assessment

Domain
performance
Issue type
Feature
Difficulty
4/5
Estimated time
3-5 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
35/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.