NVIDIA / NVIDIA/nvbench

Custom CUPTI metrics

Open
#121 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Cuda
Stars
927
Forks
123
Avg merge
2d 10h
Merged PRs (30d)
2

Description

We would like to support additional CUPTI metric types. As I understand the code right now, it seems there is a fixed set of supported metrics. As noted in #91 it's not that hard to submit a PR that adds another one, but we'd like to be able to extend this set without having to either submit a PR to nvbench, or maintain our own fork.

Would it be possible to provide an interface that can enable an arbitrary custom metric?

Contributor guide

No contributing guide indexed for this repository

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 examining how the current fixed set of CUPTI metric types is represented and extended, then review the context from issue #91. Define what an interface for arbitrary custom metrics must accept and how callers would register one; done means users can add a metric without modifying nvbench or maintaining a fork.

Written by the indexing model from the issue text.

Assessment

Domain
performance, tooling
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.