Blog
notes on kernels, chips, and everything else.
2026
- Fast Multidimensional Matrix Multiplication on CPU from Scratchloop reordering, tiling, and multithreadingJanuary 2026 · 9 min read · technical
- Pipeline Parallelismdistributed training via model partitioningFebruary 2026 · 9 min read · technical
- website refreshmy website refreshJanuary 21, 2026 · 1 min read · journal
