About This Website

Hello, thanks for your interest in my website. 😊 I try to log down my learnings and thoughts about compilers (MLIR, SPIR-V, LLVM, Triton, IREE, etc.), GPU (CUDA, ROCm, Vulkan, Metal, etc.), toolchain (ML frameworks, etc.), and others sporadically.

About Me

I’m Lei, leading both Triton teams and SHARK GPU teams to work on open-source ML compiler and runtime solutions within the AMD AI Group. Together with many other talented colleagues and community contributors, we are building Triton to be an awesome language and compiler to program AMD GPUs and IREE to be an awesome ML compiler and runtime solution to provide a smooth deployment story to meet the needs of from datacenter to edge.

Previously, I was fortunate to get to know many great Googlers and work on many great projects in Google, for SPIR-V toolchain (SPIRV-Tools, Glslang, Shaderc, DirectXShaderCompiler, rspirv, shaderc-rs), Vulkan ecosystem projects (VK-GL-CTS, GAPID, µVkCompute, hardware-perfcounter), ML toolchain (TensorFlow, TFLite, MLIR, IREE). You can find a summary here.