DeepEP

DeepEP: an efficient expert-parallel communication library

23 articles 9.4k View on GitHub ↗
23 articles
How DeepEP Optimizes Handle Reuse for Repeated Dispatches

DeepEP optimizes handle reuse for repeated dispatches by caching communication layout metadata. Improve performance by eliminating redundant calculations and GPU launches.

performance
Apr 25, 2026
DeepEP topk_idx_t Custom Dtype Requirements: Implementation Guide

Learn the requirements for DeepEP topk_idx_t custom dtype. Implement efficiently with torch.int32 or torch.int64 based on TOPK_IDX_BITS. Get the guide now.

how-to-guide
Apr 25, 2026
Does DeepEP Support Multi-Node NVLink (MNNVL)? Configuration and Usage Guide

DeepEP enables Multi-Node NVLink (MNNVL) support. Learn how to configure and use MNNVL for enhanced distributed training performance via the allow_mnnvl flag or command-line option.

how-to-guide
Apr 25, 2026
Safety Concerns with DeepEP Undefined Behavior PTX Instructions

Explore the safety concerns of DeepEP undefined behavior PTX instructions. Learn how non-standard instructions can lead to incorrect results on different GPU architectures.

deep-dive
Apr 25, 2026
How to Install and Configure the DeepEP NVSHMEM Dependency

Learn how to install and configure DeepEP NVSHMEM dependency. Follow steps for NVSHMEM v3.3.9+, set NVSHMEM_DIR, and enable GPUDirect Async for faster internode communication in DeepEP.

how-to-guide
Apr 25, 2026
DeepEP Adaptive Routing vs Static Routing: Performance Differences and Configuration Guide

Compare DeepEP adaptive routing vs static routing performance. Explore latency differences and configuration for optimal network traffic management. Discover which fits your needs.

performance
Apr 25, 2026
How DeepEP Achieves Traffic Isolation with InfiniBand Virtual Lanes

Discover how DeepEP uses InfiniBand Virtual Lanes and NVSHMEM to isolate traffic, preventing latency-sensitive MoE kernels from contending with bandwidth-heavy operations.

internals
Apr 25, 2026
How DeepEP Utilizes TMA Instructions for Minimal SM Usage on Hopper GPUs

Discover how DeepEP leverages TMA instructions for minimal SM usage on Hopper GPUs. Offloads tensor movement to TMA, achieving high bandwidth with fewer SMs.

deep-dive
Apr 25, 2026
How DeepEP Uses Zero‑Copy Optimization to Reduce SM Usage

Discover how DeepEP s zero copy optimization slashes shared memory usage by reading directly from RDMA buffers, boosting GPU occupancy and kernel efficiency. Learn more today.

deep-dive
Apr 25, 2026
DeepEP Shrink Mode for Dynamic Rank Masking: Runtime Fault Tolerance in Expert Parallelism

Explore DeepEP's shrink mode for dynamic rank masking. Temporarily disable ranks during expert-parallel communication at runtime without communicator teardown. Enhance fault tolerance.

internals
Apr 25, 2026
How to Configure DeepEP Low-Latency Mode QP (Queue Pair) for RDMA Optimization

Optimize RDMA performance by configuring DeepEP low-latency mode QP. Learn how to set num_qps_per_rank for improved buffer creation and faster communication.

how-to-guide
Apr 25, 2026
How DeepEP Enables Hook-Based Overlapping Without SM Occupation

Discover how DeepEP enables hook-based overlapping without SM occupation by launching low-latency kernels and splitting operations into distinct phases. Learn more today!

internals
Apr 25, 2026

Have a question about this repo?

These articles cover the highlights, but your codebase questions are specific. Give your agent direct access to the source. Share this with your agent to get started:

Share the following with your agent to get started:
curl -s "https://instagit.com/install.md"
Works with
Claude Codex Cursor VS Code OpenClaw Any MCP Client

Maintain an open-source project? Get it listed too →