pytorch / pytorch/cpuinfo

Add a Minimal CPU support

Open
#87 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

enhancement
Dominant language
C
Stars
1.2k
Forks
409
PR merge metrics
No merged PRs in 30d

Description

Is it possible to define/specify a minimal CPU with the very minimal set of features and a very neutral architecture ?

This specification can to be used :

  1. Provide support for not-yet-defined CPUs, or not yet existing hardware. I see that there are at least two issues , one for SPARC (#53) and the other for PPC64 (#2 ).

  2. As a test CPU. Think of fooling pytorch to use this CPU instead of the hardware based one.

  3. For profiling new code. Think of using a very sophisticated XEON while disabling every sophisticated feature and comparing the floating point results and computation times.

  4. Some post-training operations, such as deployment, explanation and prediction do not need a very sophisticated CPU.

  5. Minimal CPU is ecologically friendly.

Thanks in advance

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

Begin by reviewing the existing architecture support in cpuinfo and the related issues #53 and #2. Establish the minimal CPU feature and architecture specification needed for the listed use cases; done requires an agreed specification and corresponding implementation scope.

Written by the indexing model from the issue text.

Assessment

Tech stack
c
Domain
operating-systems
Issue type
Feature
Difficulty
5/5
Estimated time
Over a week
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
20/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.