Processing Strings 109x Faster Than Nvidia on H100(ashvardanian.com)28 points by samspenc 1 year ago | 1 comments
NumKong: 2'000 Mixed Precision Kernels for All(ashvardanian.com)39 points by ashvardanian 7 months ago | 3 comments
Python, C, Assembly – Faster Cosine Similarity(ashvardanian.com)185 points by ashvardanian 2 years ago | 45 comments
Full Unicode Search at 50× ICU Speed with AVX‑512(ashvardanian.com)189 points by ashvardanian 10 months ago | 71 comments
Matrix Core Programming on AMD CDNA3 and CDNA4 Architecture(salykova.github.io)17 points by skidrow 1 year ago | 1 comments