aliyun / aliyun/SimAI

Does the Sima code assume that the TP_num and the number of NVSwitches are consistent across all machines when initializing NCCL?

Open
#21 1 comment 0 reactions 0 assignees View on GitHub

Nobody has claimed this yet.

Dominant language
Python
Stars
1.2k
Forks
184
Avg merge
13h 4m
Merged PRs (30d)
1

Description

As shown in the figure, during initialization, MockNcclGroup calculates TP_num and NVSwitch. Does MockNcclGroup assume that TP_num and the number of NVSwitches are consistent? For example, if tp_size=4 and gpus_per_node=8, then TP_num=2. Based on the code, it seems to assume that each machine also has 2 NVSwitches in this case. Is my understanding correct? Thank you for clarifying.

image

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 MockNcclGroup's initialization and trace how it derives TP_num and NVSwitch from tp_size and gpus_per_node. Check the tp_size=4 and gpus_per_node=8 example across machines. Done means providing an evidence-based clarification of whether the values must be consistent.

Written by the indexing model from the issue text.

Assessment

Tech stack
python
Domain
distributed-systems
Issue type
Bug
Difficulty
3/5
Estimated time
1-2 days
Activity status
Stale
Clarity
Needs clarification
Newbie friendliness
25/100

Get new issues in your inbox

A short digest of beginner-friendly GitHub issues.