GEMM Optimization on Yunsheng Nihttps://niyunsheng.github.io/series/gemm-optimization/Recent content in GEMM Optimization on Yunsheng NiHugo -- 0.154.5en-usContent is licensed under CC BY-NC-SA 4.0.Sun, 22 Mar 2026 08:00:00 +0800Progressive CUDA GEMM Optimization: From Memory-Bound to Swizzlinghttps://niyunsheng.github.io/cuda-gemm-bank-conflict-swizzling/Sun, 22 Mar 2026 08:00:00 +0800https://niyunsheng.github.io/cuda-gemm-bank-conflict-swizzling/A step-by-step guide to optimizing FP32 CUDA GEMM kernels. Learn how to overcome warp-level memory coalescing bottlenecks, eliminate 32-way shared memory bank conflicts using memory padding, and implement zero-waste XOR address swizzling.Deep Dive into Triton GEMM Optimization: From Naive Tiling to Hopper TMAhttps://niyunsheng.github.io/triton-gemm-tiling-tma/Wed, 11 Feb 2026 00:39:00 +0800https://niyunsheng.github.io/triton-gemm-tiling-tma/A step-by-step guide to optimizing GEMM in Triton, covering Tiling, Autotuning, L2 Cache Optimizations, and Hopper TMA.