Blog of zenflow binding study #7614
Merged
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
There was a problem hiding this comment.
Pull Request Overview
This PR introduces a comprehensive blog/lab study examining ZenFlow technology and its performance improvements with DeepSpeed CPU core binding. The study investigates how CPU core binding affects both ZeRO Offload and ZenFlow performance, documenting specific optimizations and measurement results.
- Documents ZenFlow technology and its relationship with DeepSpeed CPU core binding
- Presents performance testing results comparing different core binding strategies
- Introduces improvements to ZenFlow's core binding mechanism developed in collaboration with ZenFlow authors
Tip: Customize your code reviews with copilot-instructions.md. Create the file or learn how to get started.
delock
force-pushed
the
gma/zenflow_binding_study
branch
2 times, most recently
from
a8a1be5 to
92d4d21
Compare
|
@delock thanks for the PR. Can you please modify https://github.com/deepspeedai/DeepSpeed/blob/master/docs/index.md in order to update the Latest News section of the home page? |
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
This PR adds a blog post for SuperOffload. More specifically, the blog covers the design and motivation behind SuperOffload, comparisons with previous approaches, key experiences and insights, and guidance on enabling and using SuperOffload. See also: [PR#7559](#7559) - SuperOffload implementation. [PR#990](deepspeedai/DeepSpeedExamples#990) - Examples. --------- Signed-off-by: Olatunji Ruwase <[email protected]> Co-authored-by: Olatunji Ruwase <[email protected]> Signed-off-by: Guokai Ma <[email protected]>
Since `make format` will generate `venv` directory, we should add it to `.gitignore`. Co-authored-by: Olatunji Ruwase <[email protected]> Signed-off-by: Guokai Ma <[email protected]>
This PR improves state management for DeepCompile in the engine. Previously, the system relied only on the config flag indicating whether DeepCompile was enabled. However, DeepCompile is actually activated only when `compile()` is called. This meant that if DeepCompile was enabled in the config but `compile()` was never called, it could lead to invalid internal states (as shown in #7598). Since `enabled == True` should be interpreted as an option that modifies the behavior of `compile()`, this PR introduces clearer state management: - If .compile() is not called, the DeepCompile config has no effect on behavior. A one-time message is shown instead. - A new state, DeepCompile activated, is introduced. This represents the condition where DeepCompile is both enabled in the config and .compile() has been called. --------- Signed-off-by: Masahiro Tanaka <[email protected]> Co-authored-by: Olatunji Ruwase <[email protected]> Signed-off-by: Guokai Ma <[email protected]>
PR #6993 replaces the flat IPG buffers with a dict maintaining type-indexed buckets. The member is also renamed from `_ipg_bucket_flat_buffer` to `ipg_buckets`. Update the bucket clearing logic in `init_z3` accordingly. Signed-off-by: Junjie Mao <[email protected]> Signed-off-by: Guokai Ma <[email protected]>
Polish SuperOffload blog post; minor grammar and style fixes --------- Co-authored-by: Olatunji Ruwase <[email protected]> Signed-off-by: Guokai Ma <[email protected]>
… when world size expansion. (#7599) When the world size expands from 2 to 4, then convert to universal checkpoint, and load from universal checkpoint. The new rank, for example, rank3 will load model file `zero_pp_rank_3_mp_rank_00_model_states.pt`. But this file was not produced during the last execution. For stage3, just load the first file, that is `zero_pp_rank_0_mp_rank_00_model_states`. The existing unit test TestZeROUniversalCheckpointDP::test_dp_world_size_2to4 can verify this problem. --------- Co-authored-by: Olatunji Ruwase <[email protected]> Signed-off-by: Guokai Ma <[email protected]>
Co-authored-by: Olatunji Ruwase <[email protected]> Signed-off-by: Guokai Ma <[email protected]>
Co-authored-by: Olatunji Ruwase <[email protected]> Signed-off-by: Guokai Ma <[email protected]>
Co-authored-by: Olatunji Ruwase <[email protected]> Signed-off-by: Guokai Ma <[email protected]>
Co-authored-by: Olatunji Ruwase <[email protected]> Signed-off-by: Guokai Ma <[email protected]>
Signed-off-by: Guokai Ma <[email protected]>
delock
force-pushed
the
gma/zenflow_binding_study
branch
from
891454d to
b21cb82
Compare
Signed-off-by: Guokai Ma <[email protected]>
|
Hi @sfc-gh-truwase the link from index page had been added, along with other wording fixes. Thanks for the suggestions! |
sfc-gh-truwase
approved these changes
Oct 6, 2025
Liangliang-Ma
pushed a commit
to Liangliang-Ma/DeepSpeed
that referenced
this pull request
Oct 13, 2025
This PR add a blog/lab for study of zenflow and zero offload performance with DeepSpeed CPU core binding. --------- Signed-off-by: Guokai Ma <[email protected]> Signed-off-by: Olatunji Ruwase <[email protected]> Signed-off-by: Masahiro Tanaka <[email protected]> Signed-off-by: Junjie Mao <[email protected]> Co-authored-by: Xinyu Lian <[email protected]> Co-authored-by: Olatunji Ruwase <[email protected]> Co-authored-by: zhengchenyu <[email protected]> Co-authored-by: Masahiro Tanaka <[email protected]> Co-authored-by: Junjie Mao <[email protected]> Signed-off-by: Ma, Liangliang <[email protected]>
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This PR add a blog/lab for study of zenflow and zero offload performance with DeepSpeed CPU core binding.