mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-05 22:22:27 +00:00

Files

Gino Lu 8ba1c708dc Add e8m0 scaled convert into CK_TILE (#2617 )

* first commit

* remove redundent code

* modify according to comments.

* fix type_convert error with scaled_type_convert

2025-08-10 07:42:42 +00:00

algorithm

[CK-TILE] File-level documentation for static encoding pattern (#2433 )

2025-07-04 02:26:18 -07:00

arch

add mixed_prec fp16xfp4

2025-08-08 20:19:16 +00:00

container

support flatmm scaling

2025-07-23 19:04:22 +00:00

numeric

Add e8m0 scaled convert into CK_TILE (#2617 )

2025-08-10 07:42:42 +00:00

tensor

Enable Async Copy for MI355 (#2425 )

2025-07-07 10:08:49 -07:00

utility

[CK_TILE] Tileloop persistent gemm - resubmit (#2299 )

2025-06-06 14:18:49 -07:00

config.hpp

default skip y point to r (#2457 )

2025-07-06 23:54:34 -07:00

README.md

introducing ck_tile! (#1216 )

2024-04-15 19:27:12 -05:00

README.md

ck_tile/core

ck_tile/core contains every basic functions and structures to create a GPU kernel using ck_tile. User should only include ck_tile/core.hpp this single header to use all the functionality. Everything is under ck_tile namespace. The coding style under this folder should be similar to std (snake_case for structure/function, Camel for template types...)

algorithm/
    coordinate transform and some other reusable algorithm
arch/
    contains some basic device building block like mma, buffer addressing, etc...
container/
    contains basic container data structure, array/sequence/tuple/...
numeric/
    data type, and data type related math
tensor/
    tensor descriptors and tile level API
utility/
    other utility function for both host/device