[https://nvbugs/5412562][feat] Allocate MoE workspace only when necessary #6502
Conversation
📝 WalkthroughWalkthroughThe code refactors workspace management in the Changes
Sequence Diagram(s)sequenceDiagram
participant Caller
participant FusedMoeRunner
Caller->>FusedMoeRunner: setWorkspaceInfo(params)
alt workspace tensor large enough
FusedMoeRunner->>FusedMoeRunner: Update src_to_dest_map pointer
else workspace tensor too small
FusedMoeRunner->>FusedMoeRunner: Log warning
FusedMoeRunner->>FusedMoeRunner: Allocate new workspace tensor
FusedMoeRunner->>FusedMoeRunner: Update src_to_dest_map pointer
end
Caller->>FusedMoeRunner: Use workspace_info member for further operations
Estimated code review effort🎯 2 (Simple) | ⏱️ ~8 minutes Suggested reviewers
Note ⚡️ Unit Test Generation is now available in beta!Learn more here, or try it out under "Finishing Touches" below. 📜 Recent review detailsConfiguration used: .coderabbit.yaml 📒 Files selected for processing (1)
🚧 Files skipped from review as they are similar to previous changes (1)
⏰ Context from checks skipped due to timeout of 90000ms. You can increase the timeout in your CodeRabbit configuration to a maximum of 15 minutes (900000ms). (1)
✨ Finishing Touches
🧪 Generate unit tests
Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. 🪧 TipsChatThere are 3 ways to chat with CodeRabbit:
SupportNeed help? Create a ticket on our support page for assistance with any issues or questions. CodeRabbit Commands (Invoked using PR comments)
Other keywords and placeholders
Documentation and Community
|
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Actionable comments posted: 1
🧹 Nitpick comments (1)
cpp/tensorrt_llm/thop/moeOp.cpp (1)
580-580: Follow member variable naming convention.The member variable
workspace_infoshould be prefixed with 'm' according to the coding guidelines:mWorkspaceInfo.- WorkspaceInfo workspace_info; + WorkspaceInfo mWorkspaceInfo;Note: This change would require updating all references to this variable throughout the class.
📜 Review details
Configuration used: .coderabbit.yaml
Review profile: CHILL
Plan: Pro
📒 Files selected for processing (1)
cpp/tensorrt_llm/thop/moeOp.cpp(5 hunks)
🧰 Additional context used
📓 Path-based instructions (2)
**/*.{cpp,h,hpp,cc,cxx}
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
**/*.{cpp,h,hpp,cc,cxx}: Closing braces of namespaces should have a comment saying the namespace it closes (e.g., } // namespace foo)
Prefer const or constexpr variables over #defines whenever possible, as the latter are not visible to the compiler.
A variable that is not modified after its initialization should be declared as const.
Except 0 (only used in comparison for checking signness/existence/emptiness) and nullptr, true, false, all other literals should only be used for variable initialization.
Use the Allman indentation style for braces.
Put the semicolon for an empty for or while loop in a new line.
The statement forming the body of a switch, while, do .. while or for statement shall be a compound statement (use brace-delimited statements).
If and else should always be followed by brace-delimited statements, even if empty or a single statement.
C++ filenames should use camel case with first letter lowercase (e.g., thisIsAFilename.cpp), and all files involved in the compilation of a target must have filenames that are case-insensitive unique.
All types (including class names) are camel case with uppercase first letter (e.g., FooBarClass).
Local variables, methods, and namespaces use camel case with first letter lowercase (e.g., localFooBar).
Non-magic-number global variables that are non-static and not defined in anonymous namespace use camel case prefixed by a lower case 'g' (e.g., gDontUseGlobalFoos).
Non-magic-number global variables that are static or defined in an anonymous namespace use camel case prefixed by a lower case 's' (e.g., sMutableStaticGlobal).
Locally visible static variable uses camel case with lowercase prefix 's' as the first letter of the name (e.g., static std::once_flag sFlag;).
Class member variables use camel case prefixed with an 'm' (e.g., mNbFooValues). Public member variables do not require the 'm' prefix but it is encouraged for clarity.
Enumerations, global constants, static constants at class-scope, and function-scope magic...
Files:
cpp/tensorrt_llm/thop/moeOp.cpp
**/*.{cpp,h,hpp,cc,cxx,cu,py}
📄 CodeRabbit Inference Engine (CODING_GUIDELINES.md)
All TensorRT-LLM Open Source Software code should contain an NVIDIA copyright header that includes the current year. This includes .cpp, .h, .cu, .py, and any other source files which are compiled or interpreted.
Files:
cpp/tensorrt_llm/thop/moeOp.cpp
🔇 Additional comments (3)
cpp/tensorrt_llm/thop/moeOp.cpp (3)
311-312: LGTM! Workspace allocation optimization implemented correctly.The calls to
setWorkspaceInfoproperly replace the previousgetWorkspaceInfopattern, supporting the PR objective of reducing frequent memory allocations. Both calling methods are already protected by mutex locks, ensuring thread safety.Also applies to: 442-443
626-647: Excellent optimization approach with minor implementation improvements needed.The
setWorkspaceInfomethod successfully implements the PR objective of reducing frequent memory allocations by maintaining workspace as a member variable and only reallocating when necessary. This should significantly improve performance by avoiding the costly cudaMalloc/cudaFree operations mentioned in the PR description.The core logic is sound and the approach is well-designed for the use case.
637-637: Assignment to int64_t is correct.Verified that
common::calculateTotalWorkspaceSizereturnssize_t. Converting its result into a signed 64-bit integer aligns with our coding guideline to prefer signed types and removes the signed/unsigned mismatch when comparing againstworkspace.numel()(which returnsint64_t). No changes are required here.
nv-yilinf
changed the title
|
/bot run |
|
PR_Github #13713 [ run ] triggered by Bot |
|
PR_Github #13713 [ run ] completed with state |
|
/bot run |
|
PR_Github #13732 [ run ] triggered by Bot |
|
PR_Github #13732 [ run ] completed with state |
|
/bot run |
|
PR_Github #13823 [ run ] triggered by Bot |
|
PR_Github #13823 [ run ] completed with state |
|
/bot run |
|
PR_Github #13829 [ run ] triggered by Bot |
|
PR_Github #13829 [ run ] completed with state |
|
/bot run |
|
PR_Github #14018 [ run ] triggered by Bot |
|
PR_Github #14018 [ run ] completed with state |
|
/bot run |
|
PR_Github #14039 [ run ] triggered by Bot |
|
PR_Github #14039 [ run ] completed with state |
nv-yilinf
changed the title
nv-yilinf
changed the title
nv-yilinf
force-pushed
the
fix-moe-workspace-allocation
branch
from
a7ea695 to
02dcf16
Compare
|
/bot run |
|
PR_Github #14172 [ run ] triggered by Bot |
|
Close this PR because of the branch retarget mess. Created a replica targeting release/1.0 #6955 |
Summary by CodeRabbit
Summary by CodeRabbit
Description
In current MoE runner implementation, the runner will allocate a new workspace Tensor each time invoking the kernel. Even though backed by torch's caching allocator, frequent cudaMalloc/Frees are usually considered not a good practice and sometimes causes cudaMalloc to take ~100ms.
In this PR we fix this issue by maintaining the workspace tensor as a Class variable and only reallocates when target size is larger than current size or when capturing cuda graph.
Test Coverage
GitHub Bot Help
/bot [-h] ['run', 'kill', 'skip', 'reuse-pipeline'] ...Provide a user friendly way for developers to interact with a Jenkins server.
Run
/bot [-h|--help]to print this help message.See details below for each supported subcommand.
run [--reuse-test (optional)pipeline-id --disable-fail-fast --skip-test --stage-list "A10-PyTorch-1, xxx" --gpu-type "A30, H100_PCIe" --test-backend "pytorch, cpp" --add-multi-gpu-test --only-multi-gpu-test --disable-multi-gpu-test --post-merge --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx" --detailed-log --debug(experimental)]Launch build/test pipelines. All previously running jobs will be killed.
--reuse-test (optional)pipeline-id(OPTIONAL) : Allow the new pipeline to reuse build artifacts and skip successful test stages from a specified pipeline or the last pipeline if no pipeline-id is indicated. If the Git commit ID has changed, this option will be always ignored. The DEFAULT behavior of the bot is to reuse build artifacts and successful test results from the last pipeline.--disable-reuse-test(OPTIONAL) : Explicitly prevent the pipeline from reusing build artifacts and skipping successful test stages from a previous pipeline. Ensure that all builds and tests are run regardless of previous successes.--disable-fail-fast(OPTIONAL) : Disable fail fast on build/tests/infra failures.--skip-test(OPTIONAL) : Skip all test stages, but still run build stages, package stages and sanity check stages. Note: Does NOT update GitHub check status.--stage-list "A10-PyTorch-1, xxx"(OPTIONAL) : Only run the specified test stages. Examples: "A10-PyTorch-1, xxx". Note: Does NOT update GitHub check status.--gpu-type "A30, H100_PCIe"(OPTIONAL) : Only run the test stages on the specified GPU types. Examples: "A30, H100_PCIe". Note: Does NOT update GitHub check status.--test-backend "pytorch, cpp"(OPTIONAL) : Skip test stages which don't match the specified backends. Only support [pytorch, cpp, tensorrt, triton]. Examples: "pytorch, cpp" (does not run test stages with tensorrt or triton backend). Note: Does NOT update GitHub pipeline status.--only-multi-gpu-test(OPTIONAL) : Only run the multi-GPU tests. Note: Does NOT update GitHub check status.--disable-multi-gpu-test(OPTIONAL) : Disable the multi-GPU tests. Note: Does NOT update GitHub check status.--add-multi-gpu-test(OPTIONAL) : Force run the multi-GPU tests in addition to running L0 pre-merge pipeline.--post-merge(OPTIONAL) : Run the L0 post-merge pipeline instead of the ordinary L0 pre-merge pipeline.--extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx"(OPTIONAL) : Run the ordinary L0 pre-merge pipeline and specified test stages. Examples: --extra-stage "H100_PCIe-TensorRT-Post-Merge-1, xxx".--detailed-log(OPTIONAL) : Enable flushing out all logs to the Jenkins console. This will significantly increase the log volume and may slow down the job.--debug(OPTIONAL) : Experimental feature. Enable access to the CI container for debugging purpose. Note: Specify exactly one stage in thestage-listparameter to access the appropriate container environment. Note: Does NOT update GitHub check status.For guidance on mapping tests to stage names, see
docs/source/reference/ci-overview.mdand the
scripts/test_to_stage_mapping.pyhelper.kill
killKill all running builds associated with pull request.
skip
skip --comment COMMENTSkip testing for latest commit on pull request.
--comment "Reason for skipping build/test"is required. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.reuse-pipeline
reuse-pipelineReuse a previous pipeline to validate current commit. This action will also kill all currently running builds associated with the pull request. IMPORTANT NOTE: This is dangerous since lack of user care and validation can cause top of tree to break.