triton-to-gluon

Community

Translate Triton kernels to Gluon for AMD GPUs.

Authorleonling-ll
Version1.0.0
Installs0

System Documentation

What problem does it solve?

Translate a Triton GPU kernel into an explicit Gluon kernel that uses AMD tensor layouts to enable efficient execution on CDNA GPUs.

Core Features & Use Cases

  • Detects GPU architecture (gfx942 for CDNA3, gfx950 for CDNA4) and selects the appropriate layout conventions.
  • Dumps Triton IR artifacts (ttgir/llir) and derives precise tensor layouts and pipeline structure.
  • Generates a Gluon kernel with explicit layouts (BlockedLayout, SwizzledSharedLayout, AMDMFMALayout, DotOperandLayout) and an integrated verify/profiler workflow.
  • Saves generated artifacts and integrates with existing porting workflows for matmul/attention/linear kernels on AMD GPUs.

Quick Start

Run the translation on your Triton kernel file, for example /triton-to-gluon <kernel_file.py>.

Dependency Matrix

Required Modules

None required

Components

Standard package

💻 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: triton-to-gluon
Download link: https://github.com/leonling-ll/claude-skills/archive/main.zip#triton-to-gluon

Please download this .zip file, extract it, and install it in the .claude/skills/ directory.
View Source Repository

Agent Skills Search Helper

Install a tiny helper to your Agent, search and equip skill from 471,000+ vetted skills library on demand.