sazc 46eb0d08fb Performance: Larger BlockTile optimizations enable 1470+ TFLOPS FP8 performance on the H800-SXM platform

2025-03-25 10:44:57 +08:00

..

include/deep_gemm

Performance: Larger BlockTile optimizations enable 1470+ TFLOPS FP8 performance on the H800-SXM platform

2025-03-25 10:44:57 +08:00

jit

Code format

2025-03-11 13:26:10 +08:00

jit_kernels

Performance: Larger BlockTile optimizations enable 1470+ TFLOPS FP8 performance on the H800-SXM platform

2025-03-25 10:44:57 +08:00

__init__.py

fix typo

2025-02-26 18:37:22 +08:00

utils.py

Correctly flush L2, as reconstructing the tensors on every iteration effectively put them in the L2, and gave the GPU enough idle time to avoid thermal throttling in a potentially unrealistic way.

2025-03-15 20:46:24 +00:00