ixccl
CommunityAccelerate distributed training with compatible NCCL library
Authordongg622
Version1.0.0
Installs0
System Documentation
What problem does it solve?
This Skill enables seamless implementation of high-performance collective communication in distributed training by providing an NCCL-compatible library optimized for TianShu chips.
Core Features & Use Cases
- Distributed Training Support: Facilitates multi-GPU and multi-node synchronization for PyTorch, DeepSpeed, and Megatron.
- Performance Tuning: Offers options to optimize communication algorithms and protocols such as Ring, Tree, LL, and LL128.
- Use Case: Enable multi-node deep learning training jobs with enhanced bandwidth and minimal latency on TianShu hardware.
Quick Start
Configure environment variables for NCCL to use ixccl, and initialize the process group in your training scripts as usual to facilitate distributed operations.
Dependency Matrix
Required Modules
None requiredComponents
scriptsreferences
💻 Claude Code Installation
Recommended: Let Claude install automatically. Simply copy and paste the text below to Claude Code.
Please help me install this Skill: Name: ixccl Download link: https://github.com/dongg622/china-ai-chip-skill/archive/main.zip#ixccl Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
Agent Skills Search Helper
Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.