Chenggang Zhao f8797d3c12 Optimize TMA issues

2025-04-14 11:29:28 +08:00

..

include/deep_gemm

Optimize TMA issues

2025-04-14 11:29:28 +08:00

jit

Compatible with padding

2025-04-11 17:38:22 +08:00

jit_kernels

Add TMA D descriptor

2025-04-11 14:10:11 +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