ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-04-20 14:59:17 +00:00
Code Issues Packages Projects Releases Wiki Activity

Optimized GEMMs for MX FP4/8 (#2294)

Browse Source 
Adds V3 GEMM pipeline for MX FP4 and MX FP8 
Adds V3 GEMM pipeline for MX FP4 with preshuffling
Adds MXFP4 GEMM tests (#2275)
Adds MXFP4 GEMM examples
Adds MXFP4 GEMMs to ckProfiler




Co-authored-by: Andriy Roshchenko <107577548+andriy-ca@users.noreply.github.com>
Co-authored-by: Andriy Roshchenko <andriy.roshchenko@amd.com>
Co-authored-by: aska-0096 <haocwang@amd.com>
Co-authored-by: lalala-sh <Jiaxing.Wen@amd.com>
Co-authored-by: OscarXu <huaiguxu@amd.com>
Co-authored-by: mtgu0705 <mtgu@amd.com>
Co-authored-by: Ding, Yi <yi.ding@amd.com>
Co-authored-by: feifei14119 <feiw@amd.com>
Co-authored-by: Lin, Qun <qlin@amd.com>
Co-authored-by: joye <joye@amd.com>
Co-authored-by: Rostyslav Geyyer <46627076+geyyer@users.noreply.github.com>
This commit is contained in:
Andriy Roshchenko
2025-06-05 13:54:15 -06:00
committed by GitHub
parent 233e274077
commit 00247e3c29
83 changed files with 8193 additions and 2165 deletions
Download Patch File Download Diff File Expand all files Collapse all files

43
include/ck/utility/functional2.hpp
View File

@@ -1,10 +1,11 @@
// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
// Copyright (c) 2018-2025, Advanced Micro Devices, Inc. All rights reserved.
#pragma once
#include "ck/utility/functional.hpp"
#include "ck/utility/sequence.hpp"
#include "ck/utility/tuple.hpp"
namespace ck {
@@ -70,4 +71,44 @@ struct static_for<0, N, 1> : detail::make_applier<N>
using detail::make_applier<N>::operator();
};
template <typename... Is>
struct static_for_range
{
template <typename F>
__host__ __device__ constexpr void operator()(F f) const
{
// tweak -fbracket-depth if compilation fails. Clang default limit is 256
(f(Is{}), ...);
}
};
template <typename... Ts>
struct static_for_product;
template <typename... Is>
struct static_for_product<Tuple<Is...>> : public static_for_range<Is...>
{
};
template <typename... Is, typename... Rest>
struct static_for_product<Tuple<Is...>, Rest...>
{
template <typename F>
__host__ __device__ constexpr void operator()(F f) const
{
static_for_product<Tuple<Is...>>{}([&](auto i0) { //
static_for_product<Rest...>{}([&](auto... is) { //
f(i0, is...);
});
});
}
};
struct identity
{
template <typename T>
__host__ __device__ constexpr T&& operator()(T&& arg) const noexcept
{
return forward<T>(arg);
}
};
} // namespace ck