Commit Graph
3 Commits
Author SHA1 Message Date
Chaofan Lin 50481cce09 [Refactor] Remove intrinsic compatibility facade (#2812) 2026-07-30 14:26:21 +08:00
Yongqi Zhuo 1cb4d4f3ac [CUDA] Support arbitrary TMEM layouts (#2785)
* [CUDA] Support arbitrary TMEM layouts

* Fix bugs
2026-07-29 13:07:28 +08:00
Tong WUandRachmanino 1fbd994cbb [Refactor] Refactor multiple TensorCoreIntrinEmitter to provide atom-level mma control interface (#2161)
* Enhance intrinsic modules by adding WGMMATensorCoreIntrinEmitter and TCGEN05TensorCoreIntrinEmitter. Introduce descriptor parameter classes for WGMMA and TCGEN05, improving the initialization and atom offset computation processes. Update the TensorCoreIntrinEmitter to support atom-level interfaces and macro definitions for efficient matrix multiplication operations.

* lint

* add test for mma/wgmma/tcgen5mma atom

* lint

* typo

* fix import path err

---------

Co-authored-by: Rachmanino <wutong.1109@bytedance.com>
2026-05-09 16:04:31 +08:00