NVIDIA / NVIDIA/cuCascade

Converters should include destination output size and temporary allocations that need to be allocated for performing allocations

Open
#16 0 comments 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

feature request improvement
Dominant language
C++
Stars
59
Forks
35
Avg merge
17h 38m
Merged PRs (30d)
3

Description

Currently converters just register a function but we also need to be able to make reservations for the destination and also be able to understand how much temporary memory we are going to need in order to perform the conversion. For example if a user wants to make a converter that will first compress the data on the gpu then we need to understand that there is going to be some gpu utilization in order to perform this conversion.

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 locating the converter registration API and the code path that performs its allocations. Trace how a conversion currently obtains destination and temporary memory, then define how reservations should represent both requirements. Done means converters can report destination output size and temporary GPU memory needs before conversion begins.

Written by the indexing model from the issue text.

Assessment

Tech stack
cpp
Domain
backend-api-design, performance
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.