--- name: perf-moe-optimization-workflow-50b430cdea7b description: Systematic workflow for MoE training optimization in Megatron Bridge, based on the Megatron-Core MoE paper. Covers the Three Walls framework, parallel folding, recompute strategy, dispatcher choice, and CUDA-graph bring-up. when_to_use: Full MoE throughput tuning sweep, or diagnosing a MoE throughput regression after a commit or config change; 'optimize MoE throughput', 'MoE perf tuning', 'Three Walls', 'memory wall', 'communication wall', 'compute wall'. --- # MoE Training Optimization Workflow Stable docs: @docs/training/moe-optimization.md Card: @skills/perf-moe-optimization-workflow/card.yaml Source: [Scalable Training of MoE Models with Megatron Core](https://arxiv.org/abs/2603.07685) ## Quick Reference Think in terms of the paper's Three Walls: - memory wall - communication wall - compute and host-overhead wall MoE tuning is iterative. Fixing one wall usually exposes the next one, so the best workflow is: fit first, scale second, profile third, then retune. ## Phase 1: Make The Run Memory-Feasible Start with a configuration that fits reliably before chasing throughput. Recommended order: 1. Use the smallest amount of model parallelism that still fits. 2. Turn on selective recompute before falling back to full recompute. 3. Add offloading only when recompute and parallelism are still insufficient. 4. Use `--fake-init-process-group` to sanity-check large parallel layouts on a single GPU before burning cluster time. ### Recompute guidance Prefer selective recompute for MoE runs: - good first choices: `layernorm`, `core_attn`, `moe_act`, `mlp`, or model-specific modules (`shared_experts`, `mla_up_proj`) - use full recompute only when the run still does not fit - revisit recompute after enabling CUDA graphs, because some graph scopes and full recompute paths do not mix well As a rule of thumb, fine-grained recompute often recovers most of the needed memory while keeping throughput much closer to the non-recompute baseline than full-layer recompute does. ## Phase 2: Choose Parallelism For Scale Priority order: 1. Maximize DP once the model fits. 2. Keep the hot communication path inside the fast interconnect when possible. 3. Use PP, plus VPP if needed, for multi-node scaling. 4. Prefer EP over extra TP for expert layers. 5. Add CP for long context once sequence length makes attention memory dominant. ### Parallel Folding Parallel Folding decouples attention and MoE parallelism so you do not have to pick a single compromise layout: ```text Attention: TP × CP × DP × PP MoE: ETP × EP × EDP × PP ``` Key knobs: - `--expert-model-parallel-size` - `--expert-tensor-parallel-size` Use it when attention prefers some TP or CP, but expert layers benefit from a larger EP degree than the dense layers can tolerate. ## Phase 3: Profile The Dominant Bottleneck | Bottleneck | What it looks like | Primary fixes | |---|---|---| | Memory | Run fits only with aggressive full recompute or OOMs during warmup | selective recompute, FP8, offloading, better PP layout | | Communication | Nsight shows large all-to-all or collective blocks | DeepEP or HybridEP, EP overlap, DP/TP overlap, better PP layout | | Host overhead | GPU gaps, launch-bound traces, Python overhead | CUDA graphs, `--manual-gc`, higher MBS, CPU affinity tuning | | Compute | Low SM utilization after comm and host issues are addressed | grouped GEMM, fusion work, FP8, dispatcher-specific kernel tuning | ## Dispatcher And Overlap Guidance Use dispatcher choice as a bottleneck fix, not as the first tuning knob. - `moe_token_dispatcher_type="alltoall"`: safest bring-up path, fine for smaller EP sizes - `moe_token_dispatcher_type="flex"` + `moe_flex_dispatcher_backend="deepep"`: strong default for H100 and B200 style deployments - `moe_token_dispatcher_type="flex"` + `moe_flex_dispatcher_backend="hybridep"`: strongest starting point on GB200 or GB300 NVL72 systems If the all-to-all path is visible in profiles, combine dispatcher tuning with: - `--overlap-moe-expert-parallel-comm` - `--overlap-grad-reduce` - `--tp-comm-overlap` ## FP8 Recipe Quick Decision | Platform | Recommended starting recipe | |---|---| | Hopper | FP8 blockwise | | Blackwell | MXFP8 | | Blackwell, speed-first exploration | NVFP4 after the BF16 or FP8 path is stable | Keep the router in FP32. The largest wins usually come from expert GEMMs and other heavy matrix math, not from trying to quantize every small MoE component. ## CUDA Graphs For MoE For dropless MoE, start with partial TE-scoped graphs: - `attn` - `moe_router` - `moe_preprocess` That path usually gives a meaningful step-time win while keeping the dynamic expert work outside the graph. Expect a moderate speedup when launch overhead is visible, but budget several extra GB of memory and verify that shapes remain static. Use full-iteration graphs only for graph-friendly workloads such as drop-and-pad or tightly controlled static-shape experiments. Related references: - @skills/perf-cuda-graphs/SKILL.md - @docs/training/cuda-graphs.md - @docs/training/activation-recomputation.md ## Pitfalls 1. **Do not optimize in the wrong order**: fitting the model and selecting sane parallelism matter more than micro-optimizations. 2. **Platform changes the limiting wall**: H100-class runs often feel more communication-bound, while GB200 or GB300 runs often expose CPU or launch overhead earlier. 3. **FP8 MFU can look misleadingly low**: compare absolute throughput as well as MFU when switching precision modes. 4. **CUDA graphs and recompute interact**: TE-scoped graphs are usually paired with selective recompute, not blanket full recompute. 5. **Parallel Folding is not optional at large scale**: once attention and expert layers want clearly different layouts, a single shared TP or EP plan becomes a tax on both. --- Source attribution: yo-steven Original project: https://github.com/yo-steven/skills-exploration-20260522/blob/6e1127ffb7690aa8038b172d0ae7cc720754870d/skills/Megatron-Bridge/perf-moe-optimization-workflow/SKILL.md License: Apache-2.0 License information: https://www.apache.org/licenses/LICENSE-2.0 Original upstream instructions; downloaded from Mega AI Hub. ## Upstream license notice SPDX-License-Identifier: Apache-2.0 AND CC-BY-4.0 Copyright (c) 2026 NVIDIA Corporation This project is licensed under a dual license: - Apache License 2.0 (for code and configuration) - Creative Commons Attribution 4.0 International (CC BY 4.0) (for documentation) You may use the contents of this repository under the terms of either license, or both, as applicable to the type of content. ================================================================================ Apache License Version 2.0, January 2004 http://www.apache.org/licenses/ TERMS AND CONDITIONS FOR USE, REPRODUCTION, AND DISTRIBUTION 1. Definitions. "License" shall mean the terms and conditions for use, reproduction, and distribution as defined by Sections 1 through 9 of this document. "Licensor" shall mean the copyright owner or entity authorized by the copyright owner that is granting the License. "Legal Entity" shall mean the union of the acting entity and all other entities that control, are controlled by, or are under common control with that entity. For the purposes of this definition, "control" means (i) the power, direct or indirect, to cause the direction or management of such entity, whether by contract or otherwise, or (ii) ownership of fifty percent (50%) or more of the outstanding shares, or (iii) beneficial ownership of such entity. "You" (or "Your") shall mean an individual or Legal Entity exercising permissions granted by this License. "Source" form shall mean the preferred form for making modifications, including but not limited to software source code, documentation source, and configuration files. "Object" form shall mean any form resulting from mechanical transformation or translation of a Source form, including but not limited to compiled object code, generated documentation, and conversions to other media types. "Work" shall mean the work of authorship, whether in Source or Object form, made available under the License, as indicated by a copyright notice that is included in or attached to the work (an example is provided in the Appendix below). "Derivative Works" shall mean any work, whether in Source or Object form, that is based on (or derived from) the Work and for which the editorial revisions, annotations, elaborations, or other modifications represent, as a whole, an original work of authorship. For the purposes of this License, Derivative Works shall not include works that remain separable from, or merely link (or bind by name) to the interfaces of, the Work and Derivative Works thereof. "Contribution" shall mean any work of authorship, including the original version of the Work and any modifications or additions to that Work or Derivative Works thereof, that is intentionally submitted to the Licensor for inclusion in the Work by the copyright owner or by an individual or Legal Entity authorized to submit on behalf of the copyright owner. For the purposes of this definition, "submitted" means any form of electronic, verbal, or written communication sent to the Licensor or its representatives, including but not limited to communication on electronic mailing lists, source code control systems, and issue tracking systems that are managed by, or on behalf of, the Licensor for the purpose of discussing and improving the Work, but excluding communication that is conspicuously marked or otherwise designated in writing by the copyright owner as "Not a Contribution." "Contributor" shall mean Licensor and any individual or Legal Entity on behalf of whom a Contribution has been received by the Licensor and subsequently incorporated within the Work. 2. Grant of Copyright License. Subject to the terms and conditions of this License, each Contributor hereby grants to You a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable copyright license to reproduce, prepare Derivative Works of, publicly display, publicly perform, sublicense, and distribute the Work and such Derivative Works in Source or Object form. 3. Grant of Patent License. Subject to the terms and conditions of this License, each Contributor hereby grants to You a perpetual, worldwide, non-exclusive, no-charge, royalty-free, irrevocable (except as stated in this section) patent license to make, have made, use, offer to sell, sell, import, and otherwise transfer the Work, where such license applies only to those patent claims licensable by such Contributor that are necessarily infringed by their Contribution(s) alone or by combination of their Contribution(s) with the Work to which such Contribution(s) was submitted. If You institute patent litigation against any entity (including a cross-claim or counterclaim in a lawsuit) alleging that the Work or a Contribution incorporated within the Work constitutes direct or contributory patent infringement, then any patent licenses granted to You under this License for that Work shall terminate as of the date such litigation is filed. 4. Redistribution. You may reproduce and distribute copies of the Work or Derivative Works thereof in any medium, with or without modifications, and in Source or Object form, provided that You meet the following conditions: (a) You must give any other recipients of the Work or Derivative Works a copy of this License; and (b) You must cause any modified files to carry prominent notices stating that You changed the files; and (c) You must retain, in the Source form of any Derivative Works that You distribute, all copyright, patent, trademark, and attribution notices from the Source form of the Work, excluding those notices that do not pertain to any part of the Derivative Works; and (d) If the Work includes a "NOTICE" text file as part of its distribution, then any Derivative Works that You distribute must include a readable copy of the attribution notices contained within such NOTICE file, excluding any notices that do not pertain to any part of the Derivative Works, in at least one of the following places: within a NOTICE text file distributed as part of the Derivative Works; within the Source form or documentation, if provided along with the Derivative Works; or, within a display generated by the Derivative Works, if and wherever such third-party notices normally appear. The contents of the NOTICE file are for informational purposes only and do not modify the License. You may add Your own attribution notices within Derivative Works that You distribute, alongside or as an addendum to the NOTICE text from the Work, provided that such additional attribution notices cannot be construed as modifying the License. You may add Your own copyright statement to Your modifications and may provide additional or different license terms and conditions for use, reproduction, or distribution of Your modifications, or for any such Derivative Works as a whole, provided Your use, reproduction, and distribution of the Work otherwise complies with the conditions stated in this License. 5. Submission of Contributions. Unless You explicitly state otherwise, any Contribution intentionally submitted for inclusion in the Work by You to the Licensor shall be under the terms and conditions of this License, without any additional terms or conditions. Notwithstanding the above, nothing herein shall supersede or modify the terms of any separate license agreement you may have executed with Licensor regarding such Contributions. 6. Trademarks. This License does not grant permission to use the trade names, trademarks, service marks, or product names of the Licensor, except as required for reasonable and customary use in describing the origin of the Work and reproducing the content of the NOTICE file. 7. Disclaimer of Warranty. Unless required by applicable law or agreed to in writing, Licensor provides the Work (and each Contributor provides its Contributions) on an "AS IS" BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied, including, without limitation, any warranties or conditions of TITLE, NON-INFRINGEMENT, MERCHANTABILITY, or FITNESS FOR A PARTICULAR PURPOSE. You are solely responsible for determining the appropriateness of using or redistributing the Work and assume any risks associated with Your exercise of permissions under this License. 8. Limitation of Liability. In no event and under no legal theory, whether in tort (including negligence), contract, or otherwise, unless required by applicable law (such as deliberate and grossly negligent acts) or agreed to in writing, shall any Contributor be liable to You for damages, including any direct, indirect, special, incidental, or consequential damages of any character arising as a result of this License or out of the use or inability to use the Work (including but not limited to damages for loss of goodwill, work stoppage, computer failure or malfunction, or any and all other commercial damages or losses), even if such Contributor has been advised of the possibility of such damages. 9. Accepting Warranty or Additional Liability. While redistributing the Work or Derivative Works thereof, You may choose to offer, and charge a fee for, acceptance of support, warranty, indemnity, or other liability obligations and/or rights consistent with this License. However, in accepting such obligations, You may act only on Your own behalf and on Your sole responsibility, not on behalf of any other Contributor, and only if You agree to indemnify, defend, and hold each Contributor harmless for any liability incurred by, or claims asserted against, such Contributor by reason of your accepting any such warranty or additional liability. END OF TERMS AND CONDITIONS ================================================================================ Creative Commons Attribution 4.0 International License (CC BY 4.0) https://creativecommons.org/licenses/by/4.0/ You are free to: Share — copy and redistribute the material in any medium or format for any purpose, even commercially. Adapt — remix, transform, and build upon the material for any purpose, even commercially. The licensor cannot revoke these freedoms as long as you follow the license terms. Under the following terms: Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use. No additional restrictions — You may not apply legal terms or technological measures that legally restrict others from doing anything the license permits. Notices: You do not have to comply with the license for elements of the material in the public domain or where your use is permitted by an applicable exception or limitation. No warranties are given. The license may not give you all of the permissions necessary for your intended use. For example, other rights such as publicity, privacy, or moral rights may limit how you use the material. For the full legal text, see: https://creativecommons.org/licenses/by/4.0/legalcode # Mega AI Hub resource metadata Creator / publisher: yo-steven Original project: https://github.com/yo-steven/skills-exploration-20260522/blob/6e1127ffb7690aa8038b172d0ae7cc720754870d/skills/Megatron-Bridge/perf-moe-optimization-workflow/SKILL.md Attribution basis: Source repository record Fields of work: Data & Analytics, Development & Engineering Roles: Backend Developer, Data Scientist, ML / AI Engineer Specialties: Backend Tools: Megatron Country applicability: Country not specified Classification basis: Source catalog labels Instruction SHA-256: 1f83db7e7308af0537c03fe9160aec012740c67f8d96cd0dbed22d25c135a966