Han Guo Plans Study of Blackwell Kernels

MIT CSAIL PhD student Han Guo posted on X that it is time to learn more about Blackwell kernels. He linked to Luke D. Huang’s X post, which in turn links to a blog on writing speed-of-light GEMM kernels for the NVIDIA B200. The post covers GPU architecture, CuTeDSL fundamentals, loop tiling, TMA, TMEM, and swizzling. The single post in the visible conversation received two likes and four bookmarks. No additional outcomes or timing details appear in the evidence.