3 people 2cccf1f5dd [Feature] Upgrade cutlass version and support fp8 T.gemm (#202) ...

* upgrade cutlass to upstream v3.8.0

* Implement fp8 gemm and add example script

* Fix dtype retrieval with map_torch_type for fp8 inputs

* Disable vectorization of fp8 values

* Make MMA declaration compatible with cutlass 3.4.0+

* Add test for fp8 T.gemm

* fix indent

* fix indent

* Add copyright and license header

* Add copyright and license header

* lint fix

* Refactor matmul_nt and assert_matmul_correctness functions for improved readability by consolidating parameter definitions and adjusting formatting.

* clang format lint

---------

Co-authored-by: Lei Wang <34334180+LeiWang1999@users.noreply.github.com>
Co-authored-by: LeiWang1999 <leiwang1999@outlook.com>

2025-03-13 19:51:03 +08:00

..

amd

[Refactor] Simplify interface via replacing argument thread binding of intrinsics with KernelFrame.Current (#34)

2025-01-23 21:03:10 +08:00

autotune

[Dev][Benchmark] Add MLA paged decoding example and benchmark script (#158)

2025-03-06 17:40:50 +08:00

cache

[Feat] Introduce new caching mechanism for compiled kernels (#176)

2025-03-09 20:29:37 +08:00

carver

[CostModel][Carver] Support Hint Recommend for Shared memory Kernel Fusion (#73)

2025-02-11 19:32:59 +08:00

cpu

Add cpu jit with backend ctypes (#154)

2025-03-06 22:26:42 +08:00

debug

[Enhancement] Support debug print for unsigned char datatype (#145)

2025-03-05 18:29:52 +08:00

dynamic

[JIT] Enhance cython/ctypes wrapper for tma descriptor (#126)

2025-02-27 23:15:14 +08:00

ir

[Initialization] Migration of Codebase from Dev Branch into Main (#10)

2025-01-11 17:03:47 +08:00

issue

[JIT] Support Cython jit and make cython a default execution backend (#102)

2025-02-21 22:44:44 +08:00

jit

[Bugfix] Add dynamic shape support with out_idx in Cython JIT kernel compilation (#185)

2025-03-11 15:08:16 +08:00

kernel

[Feature] Upgrade cutlass version and support fp8 T.gemm (#202)

2025-03-13 19:51:03 +08:00

language

[Language] Support clamp in language (#192)

2025-03-12 16:29:18 +08:00

primitives

[Tools] Introduce plot_layout to visualize the fragment layout (#68)

2025-02-10 01:03:01 +08:00

tilelibrary

[Doc] Remove unnecessary layout annotation (#49)

2025-01-25 17:05:05 +08:00

transform

[Feature] Add TMA Store Synchronization Support (#195)

2025-03-12 20:06:23 +08:00

webgpu

[Backend][WebGPU] Support WebGPU WGSL code generation (#86)

2025-02-15 18:28:29 +08:00