Triton-distributed

Getting Started

Installation
Tutorials
End-to-End Integration
Megakernel Implementations
Intra-Kernel Profiler User Guide
Build Triton-distributed
Running Tests
Primitives Provided by Triton-distributed
AutoTuner for Triton-distributed

Kernels & Layers

Kernels
- NVIDIA Kernels
- AMD Kernels
  - Kernel List
Layers
Models

Python API

triton-dist.language
Triton-distributed Semantics

LittleKernel

LittleKernel

Advanced Topics

Performance of Triton-distributed on AMD GPUs
Download and fix NVSHMEM
End-to-End Demo for Triton-Distributed
How to pull upstream code

Examples

Tutorials

Triton-distributed Documents
Triton-distributed 自动调优器

Triton-distributed

Kernels
AMD Kernels
GEMM AllReduce (AMD)
View page source

GEMM AllReduce (AMD)

GEMM + AllReduce kernel for AMD GPUs.

See the gemm_allreduce.py module in python/triton_dist/kernels/amd/ for implementation details.

Previous Next

© Copyright 2025 ByteDance Seed Infrastructures Team.

Built with Sphinx using a theme provided by Read the Docs.