TIRx Harness: An Open Compiler Harness for Agentic GPU Programming

Making AMD GPUs competitive for LLM inference (2023)

Making AMD GPUs competitive for LLM inference

GPU-Accelerated LLM on an Orange Pi