nemo-mbridge-perf-moe-comm-overlap

nemo-mbridge-perf-moe-comm-overlap is a skill for Claude Code from NVIDIA-NeMo/Megatron-Bridge. It costs 42 tokens per session (1,929 once invoked), scanned A, a copy of nemo-mbridge-perf-moe-comm-overlap, Apache-2.0.

A guide to overlapping mixture-of-experts communication with computation in Megatron-Bridge. Mixture-of-experts models route tokens to specialist sub-networks, and this guide covers overlapping that routing and return traffic with other work.

In plain words
What is it for?
Enabling and validating expert-parallel communication overlap, configuring token dispatchers, scheduling expert weight-gradient work, and tuning training throughput.
Why use it?
It can help reduce time spent waiting for token transfers when communication is visible in performance profiles. The guide also identifies prerequisites and settings that can affect correctness.

Skill for Claude Code

Written for Claude Code: when-to-use in frontmatter.

Needs its repository: it runs a file that does not travel with it, so clone the repository first. The line is uv run python scripts/performance/run_script.py \.

Good fit Enabling and validating expert-parallel communication overlap, configuring token dispatchers, scheduling expert weight-gradient work, and tuning training throughput.

Compare 6 skills from other repositories ↓
Install

Getting it into your agent

It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.

Clone the repo
git clone --depth 1 https://github.com/NVIDIA-NeMo/Megatron-Bridge
agentmods
npx agentmods add skills/nvidia-nemo/megatron-bridge/nemo-mbridge-perf-moe-comm-overlap

Made for: Claude Code.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for nemo-mbridge-perf-moe-comm-overlap

README.md
[![agentmods](https://agentmods.dev/badge/skills/nvidia-nemo/megatron-bridge/nemo-mbridge-perf-moe-comm-overlap/github.svg)](https://agentmods.dev/skills/nvidia-nemo/megatron-bridge/nemo-mbridge-perf-moe-comm-overlap)
Your own site
<a href="https://agentmods.dev/skills/nvidia-nemo/megatron-bridge/nemo-mbridge-perf-moe-comm-overlap"><img src="https://agentmods.dev/badge/skills/nvidia-nemo/megatron-bridge/nemo-mbridge-perf-moe-comm-overlap/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for nemo-mbridge-perf-moe-comm-overlap

Your own site · 80×15
<a href="https://agentmods.dev/skills/nvidia-nemo/megatron-bridge/nemo-mbridge-perf-moe-comm-overlap"><img src="https://agentmods.dev/badge/skills/nvidia-nemo/megatron-bridge/nemo-mbridge-perf-moe-comm-overlap.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 42 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,929 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe. Third-party audits
  • NVIDIA SkillSpector pass 7 Sept 2026
How audits are shown
Origin 100% copy Near-identical to another mod in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00042 $0.01929
Opus 5 $0.00021 $0.00964
Sonnet 5 $0.00008 $0.00386
Haiku 4.5 $0.00004 $0.00193

Measured 12d ago against content hash 4a9656bdc898, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-11, from the pricing page.

Security

Grade A, and why

nemo-mbridge-perf-moe-comm-overlap scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 12d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

Origin

This is a copy

100% identical to nemo-mbridge-perf-moe-comm-overlap — 0 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.

skills/nemo-mbridge-perf-moe-comm-overlap/SKILL.md · 189 lines

How it starts

The opening of the file, as written. The whole thing — 189 lines — stays where its author put it; the contents beside it link to each section on GitHub.

MoE Communication Overlap

For the higher-level overview, see:

  • @docs/training/communication-overlap.md
  • @skills/nemo-mbridge-perf-moe-comm-overlap/card.yaml

Quick Decision

Use MoE communication overlap when:

  • EP > 1
  • token dispatch or combine time is visible in the profile
  • the run is already correct and you are now tuning throughput

Avoid turning it on as an early bring-up step. It is easier to validate after the dispatcher, routing mode, and recompute plan are already stable.

Enablement

cfg.comm_overlap.overlap_moe_expert_parallel_comm = True

# Optional: delayed wgrad for additional overlap
cfg.comm_overlap.delay_wgrad_compute = True

# IMPORTANT: disable shared expert overlap when using dispatch overlap
cfg.model.moe_shared_expert_overlap = False

Prerequisites

  • expert_model_parallel_size > 1
  • num_moe_experts > 1
  • moe_token_dispatcher_type must be "alltoall" or "flex"
  • Precision: BF16 or FP16
  • If PP is used, VPP (virtual_pipeline_model_parallel_size) must be set (non-None)

Flex dispatcher activation

Setting moe_flex_dispatcher_backend alone does not activate flex dispatch. You must also set moe_token_dispatcher_type = "flex".

Recompute And CUDA Graph Interaction

  • Full recompute is not a good companion for the overlap path.
  • delay_wgrad_compute adds further constraints if CUDA-graph scopes include attention or MoE-router work.
  • In practice, selective recompute is the safer pairing when overlap is enabled.

Measured Evidence

HybridEP production-shape validation

A 2026-07-25 controlled Qwen3 30B-A3B pretraining comparison used 16 H100 GPUs, BF16, sequence length 4096, TP=1, PP=1, CP=1, EP=16, MBS=1, GBS=1024, forced-balanced routing, HybridEP, and Transformer Engine CUDA-graph scopes moe_router and moe_preprocess. The only performance change was plain EP overlap; delayed wgrad stayed disabled.

Case Steady window Step time Model TFLOPS/GPU
EP overlap off iterations 5-20 24.7138s 244.039
EP overlap on, search run iterations 5-20 21.0725s 286.208
EP overlap on, independent validation iterations 41-50 20.9920s 287.305

Read the full file on GitHub · 189 lines

Files

What ships with it

5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 12d ago First seen · 189 lines · 42 tokens per session scan A 4a9656bdc898

Subscribe to this mod's changes

nemo-mbridge-perf-moe-comm-overlap is a skill published in the GitHub repository NVIDIA-NeMo/Megatron-Bridge (905 stars, last pushed yesterday), licensed Apache-2.0. It adds 42 tokens to every session and 1,929 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. It is 100% identical to nemo-mbridge-perf-moe-comm-overlap, differing in 0 lines, and is treated as a copy.

Related

Other skills, from other repositories

agent-platform-rag-engine-management

Manage and query Agent Platform RAG Engine Corpora and retrieve grounded contexts using the Google GenAI SDK. Use when listing RAG corpora or files, inspecting a corpus, retrieving contexts, or generating content grounded in a RAG corpus. Do not use for standard database queries (use SQL/Spanner skills), Google…

google/skills · 85 tokens

agent-platform-model-registry

Agent Platform Model Registry Management. Use when you need to upload, list, describe, update, or delete machine learning models (and their versions) in the Agent Platform Model Registry. Don't use for model training, model deployment to endpoints, or managing non-Agent Platform models.

google/skills · 60 tokens

foundry-config-setup

Resolve missing setup caused by a hardcoded Foundry project endpoint or model in a sample. Use when a sample fails because it uses a placeholder/hardcoded projectendpoint (for example "https://your-project.services.ai.azure.com") or a hardcoded model instead of reading them from the environment.

microsoft/agent-framework · 65 tokens

google-cloud-solution-agentic-analytics-spark-knowledge-catalog

Discovers requirements and generates guidance to design and deploy a governed, secure agentic-analytics solution for data that's distributed across Google Cloud, other cloud providers, or on-premises. Data that's outside Google Cloud (such as data from Databricks, Snowflake, Salesforce, SAP, or Oracle systems) is…

google/skills · 138 tokens

training-check

Interactively monitor training metrics from the current Codex session, periodically checking WandB or fallback logs for NaN, divergence, plateaus, and broken runs.

wanshuiyin/Auto-claude-code-research-in-sleep · 35 tokens

nemo-automodel-launcher-config

Configure NeMo AutoModel job launches for interactive runs, Slurm clusters, and SkyPilot cloud execution.

NVIDIA/skills · 30 tokens