ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-05-04 05:31:24 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add OCP FP8 support in CK_TILE (#1829)

Browse Source 
* Add OCP FP8 to CK_TILE

* Validate OCP FP8 in FMHA FWD under VALID=1
This commit is contained in:
Andriy Roshchenko
2025-01-27 11:59:49 -07:00
committed by GitHub
parent 39dc25a9b8
commit 35aebe5936
5 changed files with 767 additions and 495 deletions
Download Patch File Download Diff File Expand all files Collapse all files

18
include/ck_tile/core/config.hpp
View File

@@ -1,13 +1,13 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved.
#pragma once
#if defined(__gfx908__) || defined(__gfx90a__) || defined(__gfx940__) || defined(__gfx941__) || \
defined(__gfx942__)
defined(__gfx942__) || defined(__gfx950__)
#define __gfx9__
#endif
#if defined(__gfx940__) || defined(__gfx941__) || defined(__gfx942__)
#if defined(__gfx940__) || defined(__gfx941__) || defined(__gfx942__) || defined(__gfx950__)
#define __gfx94__
#endif
#if defined(__gfx1030__) || defined(__gfx1031__) || defined(__gfx1032__) || \
@@ -230,3 +230,15 @@
#ifndef CK_TILE_REFERENCE_MOE_SORTING_MOCK_ID
#define CK_TILE_REFERENCE_MOE_SORTING_MOCK_ID 1
#endif
#ifndef __HIP_DEVICE_COMPILE__ // for host code
#ifdef CK_TILE_USE_OCP_FP8
#define CK_TILE_USE_OCP_FP8 1
#else
#define CK_TILE_USE_OCP_FP8 0
#endif
#elif defined(__gfx950__) || defined(__gfx12__) // for GPU code
#define CK_TILE_USE_OCP_FP8 1
#else // for GPU code
#define CK_TILE_USE_OCP_FP8 0
#endif