ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-06-29 11:16:59 +00:00
Code Issues Packages Projects Releases Wiki Activity
Files
new_develop
composable_kernel/experimental/gemm_benchmark/common.hpp
Illia Silin 717f2efef7 [rocm-libraries] ROCm/rocm-libraries#6978 (commit e58096d) 
[CK] add composable kernel support on gfx1250 (#6978)

## Motivation

Add composable kernel support on gfx1250.

## Technical Details

<!-- Explain the changes along with any relevant GitHub links. -->

## Test Plan

<!-- Explain any relevant testing done to verify this PR. -->

## Test Result

<!-- Briefly summarize test outcomes. -->

## Submission Checklist

- [ ] Look over the contributing guidelines at
https://github.com/ROCm/ROCm/blob/develop/CONTRIBUTING.md#pull-requests.

---------

Co-authored-by: Qun Lin <qlin@amd.com>
Co-authored-by: jialuo12_amdeng <jia.luo@amd.com>
Co-authored-by: Andriy Roshchenko <andriy.roshchenko@amd.com>
Co-authored-by: hsivasun_amdeng <haresh.sivasuntharampillai@amd.com>
2026-05-15 06:46:51 -07:00

739 lines
22 KiB
C++
Raw Permalink Blame History

// SPDX-License-Identifier: MIT
// Copyright (c) 2018-2024, Advanced Micro Devices, Inc. All rights reserved.
#pragma once
#include <cstdlib>
#include <iostream>
#include <initializer_list>
#include <numeric>
#include <unordered_map>
#include "ck/ck.hpp"
#include "ck/tensor_operation/gpu/device/gemm_specialization.hpp"
#include "ck/tensor_operation/gpu/device/tensor_layout.hpp"
#include "ck/tensor_operation/gpu/element/element_wise_operation.hpp"
#include "ck/utility/data_type.hpp"
#include "ck/tensor_operation/gpu/grid/block_to_ctile_map.hpp"
#include "ck/library/utility/check_err.hpp"
#include "ck/library/utility/device_memory.hpp"
#include "ck/library/utility/fill.hpp"
#include "ck/library/utility/host_tensor.hpp"
#include "ck/library/utility/host_tensor_generator.hpp"
#include "ck/library/utility/literals.hpp"
#include "ck/library/reference_tensor_operation/cpu/reference_gemm.hpp"
#include "ck/library/reference_tensor_operation/gpu/reference_gemm.hpp"
using ::ck::DeviceMem;
using ::ck::HostTensorDescriptor;
using ::ck::Tensor;
struct ProblemSize final
{
ck::index_t M = 3840;
ck::index_t N = 4096;
ck::index_t K = 4096;
ck::index_t StrideA = -1;
ck::index_t StrideB = -1;
ck::index_t StrideC = -1;
};
struct ProblemSizeStreamK final
{
ck::index_t M = 3840;
ck::index_t N = 4096;
ck::index_t K = 4096;
ck::index_t StrideA = -1;
ck::index_t StrideB = -1;
ck::index_t StrideC = -1;
ck::index_t NumSKBlocks = -1; // number of stream-k blocks
};
struct ProblemSizeStreamK_universal final
{
ck::index_t M = 3840;
ck::index_t N = 4096;
ck::index_t K = 4096;
ck::index_t StrideA = -1;
ck::index_t StrideB = -1;
ck::index_t StrideC = -1;
ck::index_t Grid_size = -1; // defaults to max occupancy
ck::index_t Streamk_sel = 1; // defaults to 1-tile SK
ck::StreamKReductionStrategy reduction_strategy = ck::StreamKReductionStrategy::Atomic;
};
struct ProblemSizeSplitK final
{
ck::index_t M = 3840;
ck::index_t N = 4096;
ck::index_t K = 4096;
ck::index_t StrideA = -1;
ck::index_t StrideB = -1;
ck::index_t StrideC = -1;
ck::index_t KBatch = 1;
};
struct ExecutionConfig final
{
// 0 - no verification, 1 - CPU, 2 - GPU, 3 - CPU + GPU
int do_verification = 1;
int init_method = 2;
bool time_kernel = false;
int instance_index = -1;
int cold_niters = 50;
int nrepeat = 100;
int rotating_count = 4;
int verbosity = 1;
};
template <ck::index_t... Is>
using S = ck::Sequence<Is...>;
using Row = ck::tensor_layout::gemm::RowMajor;
using Col = ck::tensor_layout::gemm::ColumnMajor;
using PassThrough = ck::tensor_operation::element_wise::PassThrough;
template <typename ProblemType>
bool parse_cmd_args(int, char*[], ProblemType&, ExecutionConfig&)
{
return false;
}
template <>
bool parse_cmd_args<ProblemSize>(int argc,
char* argv[],
ProblemSize& problem_size,
ExecutionConfig& config)
{
if(argc == 1)
{
// use default case
}
else if(argc == 4)
{
config.do_verification = std::stoi(argv[1]);
config.init_method = std::stoi(argv[2]);
config.time_kernel = std::stoi(argv[3]);
}
else if(argc == 10)
{
config.do_verification = std::stoi(argv[1]);
config.init_method = std::stoi(argv[2]);
config.time_kernel = std::stoi(argv[3]);
problem_size.M = std::stoi(argv[4]);
problem_size.N = std::stoi(argv[5]);
problem_size.K = std::stoi(argv[6]);
problem_size.StrideA = std::stoi(argv[7]);
problem_size.StrideB = std::stoi(argv[8]);
problem_size.StrideC = std::stoi(argv[9]);
}
else
{
std::cerr
<< "arg1: verification (0=no, 1=CPU, 2=GPU, 3=CPU and GPU)" << std::endl
<< "arg2: initialization (0=no init, 1=integer value, 2=decimal value)" << std::endl
<< "arg3: time kernel (0=no, 1=yes)" << std::endl
<< "arg4 to 9: M (256x), N(128x), K(32x), StrideA, StrideB, StrideC (default: -1 or 0)"
<< std::endl;
return false;
}
return true;
}
template <>
bool parse_cmd_args<ProblemSizeStreamK_universal>(int argc,
char* argv[],
ProblemSizeStreamK_universal& problem_size,
ExecutionConfig& config)
{
if(argc == 1)
{
// use default case
}
else if(argc == 4)
{
config.do_verification = std::stoi(argv[1]);
config.init_method = std::stoi(argv[2]);
config.time_kernel = std::stoi(argv[3]);
}
else if(argc >= 10)
{
config.do_verification = std::stoi(argv[1]);
config.init_method = std::stoi(argv[2]);
config.time_kernel = std::stoi(argv[3]);
problem_size.M = std::stoi(argv[4]);
problem_size.N = std::stoi(argv[5]);
problem_size.K = std::stoi(argv[6]);
problem_size.StrideA = std::stoi(argv[7]);
problem_size.StrideB = std::stoi(argv[8]);
problem_size.StrideC = std::stoi(argv[9]);
if(argc >= 11)
{
problem_size.Streamk_sel = std::stoi(argv[10]);
if(argc >= 12)
{
problem_size.Grid_size = std::stoi(argv[11]);
if(argc >= 13)
{
int reduction_strategy = std::stoi(argv[12]);
problem_size.reduction_strategy = reduction_strategy == 0
? ck::StreamKReductionStrategy::Atomic
: ck::StreamKReductionStrategy::Reduction;
}
}
}
}
else
{
std::cerr
<< "arg1: verification (0=no, 1=CPU, 2=GPU, 3=CPU and GPU)" << std::endl
<< "arg2: initialization (0=no init, 1=integer value, 2=decimal value)" << std::endl
<< "arg3: time kernel (0=no, 1=yes)" << std::endl
<< "arg4 to 9: M (256x), N(128x), K(32x), StrideA, StrideB, StrideC (default: -1 or 0)"
<< std::endl
<< "arg10: stream-k select (-1: default config, 0: all DP, 1: 1-tile SK, 2: 2-tile SK)"
<< std::endl
<< "arg11: Grid_size(-1 for max occupancy)" << std::endl
<< "arg12: Reduction strategy (0: Atomic, 1: Reduction)" << std::endl;
return false;
}
return true;
}
template <>
bool parse_cmd_args<ProblemSizeStreamK>(int argc,
char* argv[],
ProblemSizeStreamK& problem_size,
ExecutionConfig& config)
{
if(argc == 1)
{
// use default case
}
else if(argc == 4)
{
config.do_verification = std::stoi(argv[1]);
config.init_method = std::stoi(argv[2]);
config.time_kernel = std::stoi(argv[3]);
}
else if(argc >= 10)
{
config.do_verification = std::stoi(argv[1]);
config.init_method = std::stoi(argv[2]);
config.time_kernel = std::stoi(argv[3]);
problem_size.M = std::stoi(argv[4]);
problem_size.N = std::stoi(argv[5]);
problem_size.K = std::stoi(argv[6]);
problem_size.StrideA = std::stoi(argv[7]);
problem_size.StrideB = std::stoi(argv[8]);
problem_size.StrideC = std::stoi(argv[9]);
if(argc >= 11)
{
problem_size.NumSKBlocks = std::stoi(argv[10]);
}
}
else
{
std::cerr
<< "arg1: verification (0=no, 1=CPU, 2=GPU, 3=CPU and GPU)" << std::endl
<< "arg2: initialization (0=no init, 1=integer value, 2=decimal value)" << std::endl
<< "arg3: time kernel (0=no, 1=yes)" << std::endl
<< "arg4 to 9: M (256x), N(128x), K(32x), StrideA, StrideB, StrideC (default: -1 or 0)"
<< std::endl
<< "arg10: stream-k select (0: all DP, 1: 1-tile SK, 2: 2-tile SK)"
<< "\narg11: Grid_size(-1 for max occupancy)" << std::endl;
return false;
}
return true;
}
template <>
bool parse_cmd_args<ProblemSizeSplitK>(int argc,
char* argv[],
ProblemSizeSplitK& problem_size,
ExecutionConfig& config)
{
if(argc == 1)
{
// use default case
}
else if(argc == 4)
{
config.do_verification = std::stoi(argv[1]);
config.init_method = std::stoi(argv[2]);
config.time_kernel = std::stoi(argv[3]);
}
else if(argc >= 10)
{
config.do_verification = std::stoi(argv[1]);
config.init_method = std::stoi(argv[2]);
config.time_kernel = std::stoi(argv[3]);
problem_size.M = std::stoi(argv[4]);
problem_size.N = std::stoi(argv[5]);
problem_size.K = std::stoi(argv[6]);
problem_size.StrideA = std::stoi(argv[7]);
problem_size.StrideB = std::stoi(argv[8]);
problem_size.StrideC = std::stoi(argv[9]);
if(argc >= 11)
{
problem_size.KBatch = std::stoi(argv[10]);
}
if(argc >= 12)
{
config.instance_index = std::stoi(argv[11]);
}
if(argc >= 13)
{
config.cold_niters = std::stoi(argv[12]);
}
if(argc >= 14)
{
config.nrepeat = std::stoi(argv[13]);
}
if(argc >= 15)
{
config.rotating_count = std::stoi(argv[14]);
}
}
else
{
std::cerr
<< "arg1: verification (0=no, 1=CPU, 2=GPU, 3=CPU and GPU)" << std::endl
<< "arg2: initialization (0=no init, 1=integer value, 2=decimal value)" << std::endl
<< "arg3: time kernel (0=no, 1=yes)" << std::endl
<< "arg4 to 9: M (256x), N(128x), K(32x), StrideA, StrideB, StrideC (default: -1 or 0)"
<< std::endl
<< "arg10: KBatch" << std::endl
<< "arg11-14(optional): instance_index warmup_iters repeat_iters rotating_count"
<< std::endl;
return false;
}
return true;
}
template <typename DataType, typename ComputeDataType = DataType>
inline __host__ __device__ constexpr double get_rtol()
{
if constexpr(std::is_same_v<DataType, float> && std::is_same_v<ComputeDataType, ck::tf32_t>)
{
return 1e-3;
}
else if constexpr(std::is_same_v<DataType, float>)
{
return 1e-3;
}
else if constexpr(std::is_same_v<DataType, double>)
{
return 1e-6;
}
else if constexpr(std::is_same_v<DataType, ck::half_t>)
{
return 1e-3;
}
else if constexpr(std::is_same_v<DataType, ck::bhalf_t>)
{
return 5e-2;
}
else if constexpr(std::is_same_v<DataType, int32_t>)
{
return 1e-1;
}
else if constexpr(std::is_same_v<DataType, int8_t>)
{
return 1e-1;
}
else if constexpr(std::is_same_v<DataType, ck::f8_t>)
{
return 1e-1; // 240 and 224 are acceptable
}
else if constexpr(std::is_same_v<DataType, ck::bf8_t>)
{
return 1.5e-1; // 57344 and 49152 are acceptable
}
else
{
return 1e-3;
}
}
template <typename DataType, typename ComputeDataType = DataType>
inline __host__ __device__ constexpr double get_atol()
{
if constexpr(std::is_same_v<DataType, float> && std::is_same_v<ComputeDataType, ck::tf32_t>)
{
return 1e-3;
}
else if constexpr(std::is_same_v<DataType, float>)
{
return 1e-3;
}
else if constexpr(std::is_same_v<DataType, double>)
{
return 1e-6;
}
else if constexpr(std::is_same_v<DataType, ck::half_t>)
{
return 1e-3;
}
else if constexpr(std::is_same_v<DataType, ck::bhalf_t>)
{
return 5e-2;
}
else if constexpr(std::is_same_v<DataType, int32_t>)
{
return 1e-1;
}
else if constexpr(std::is_same_v<DataType, int8_t>)
{
return 1e-1;
}
else if constexpr(std::is_same_v<DataType, ck::f8_t>)
{
return 16.1; // 240 and 224 are acceptable
}
else if constexpr(std::is_same_v<DataType, ck::bf8_t>)
{
return 8192.1; // 57344 and 49152 are acceptable
}
else
{
return 1e-3;
}
}
template <bool KLast>
void preShuffleScaleBuffer_gfx950(ck::e8m0_bexp_t* src, ck::e8m0_bexp_t* dst, int MN, int K)
{
int MNXdlPack = 2;
int KXdlPack = 2;
int XdlMNThread = 16;
int XdlKThread = 64 / XdlMNThread;
int K0 = K / KXdlPack / XdlKThread; // KRepeat
// On gfx950, WarpSize=64:
// The 4 16x128 building blocks will be packed into 1 32x256
// The 8 16x16x128 mfma will be packed into 1 32x32x256
// unfold the MN32xK(256/32) scale buffer
// 4 16 2 2
// To XdlKThread-> XdlMNThread -> KXdlPack -> MNXdlPack
// Then, MNRepeat->KRepeat
for(int n = 0; n < MN; ++n)
{
for(int k = 0; k < K; ++k)
{
int n0 = n / (XdlMNThread * MNXdlPack); // i MNRepeat
int tempn = n % (XdlMNThread * MNXdlPack);
int n1 = tempn % XdlMNThread; // i XdlMNThread
int n2 = tempn / XdlMNThread; // i MNXdlPack
int k0 = k / (XdlKThread * KXdlPack); // i KRepeat
int tempk = k % (XdlKThread * KXdlPack);
int k1 = tempk % XdlKThread; // i XdlKThread
int k2 = tempk / XdlKThread; // i KXdlPack
int outputIndex = n0 * MNXdlPack * KXdlPack * XdlMNThread * XdlKThread * K0 +
k0 * MNXdlPack * KXdlPack * XdlMNThread * XdlKThread +
k1 * MNXdlPack * KXdlPack * XdlMNThread + n1 * MNXdlPack * KXdlPack +
k2 * MNXdlPack + n2;
// src[n * K + k] = ck::type_convert<ck::e8m0_bexp_t>(static_cast<float>(powf(2.0f,
// 2-k)));
if constexpr(KLast)
dst[outputIndex] = src[n * K + k];
else
dst[outputIndex] = src[k * MN + n];
}
}
}
/**
* Pre-shuffle scale buffer for gfx1250 16x16x128 wmma scale instruction
*
* @tparam ScaleType Scale data type
* @tparam KStride Whether K is the leading dimension of the scale buffer
*/
template <typename ScaleType, ck::index_t ScaleBlockSize, bool KStride>
void preShuffleScaleBuffer_gfx1250(const ScaleType* src,
ScaleType* dst,
ck::index_t MN,
ck::index_t K)
{
static_assert(ScaleBlockSize == 32 && sizeof(ScaleType) == 1,
"wrong! only support 8-bit scale with ScaleBlockSize=32");
constexpr ck::index_t MPerXdlops = 16;
// constexpr ck::index_t NPerXdlops = 16;
constexpr ck::index_t KPerXdlops = 128;
int MNPack = 2; // 2 sets of scales in M/N direction
int KPack = 1; // 1 set of scales in K direction
int MNStep = MPerXdlops;
int KStep = KPerXdlops / ScaleBlockSize; // scales per thread
int K0 = K / KPack / KStep; // KRepeat - how many KStep blocks
// On gfx1250, WarpSize=32:
// -- The 2 16x128 building blocks will be packed into 1 32x128
// -- The 4 16x16x128 wmma will be packed into 1 32x32x128
// unfold the MN32xK(128/32) scale buffer
// 4 16 1 2
// To KStep -> MNStep -> KPack -> MNPack
// or ???
// 2 16 1 4
// MNPack -> MNStep -> KPack -> KStep
for(int mn = 0; mn < MN; ++mn)
{
int iMNRepeat = mn / (MNStep * MNPack); // i MNRepeat (MN block id)
int tempmn = mn % (MNStep * MNPack); // position in MN block
for(int k = 0; k < K; ++k)
{
int iKRepeat = k / (KStep * KPack); // i KRepeat
int tempk = k % (KStep * KPack); // position in KStep block
int outputIndex = (iMNRepeat * MNPack * MNStep) * (KStep * KPack * K0) +
(iKRepeat * KStep * KPack) * (MNStep * MNPack) +
tempmn * (KStep * KPack) + tempk;
if constexpr(KStride)
{
dst[outputIndex] = src[mn * K + k];
}
else
dst[outputIndex] = src[k * MN + mn];
}
}
}
template <typename T>
void preShuffleBuffer(const T* src, T* dst, int N, int K, int NXdl, int KPack)
{
int NLane = NXdl;
int KLane = ck::get_warp_size() / NLane;
int K_pk = std::is_same_v<T, ck::f4x2_pk_t> ? K / 2 : K;
int K0 = K_pk / (KLane * KPack);
// K -> K0 KLane KPack
// N -> N0 NLane
// N, K -> N0 K0 KLane NLane KPack
int tempk;
for(int n = 0; n < N; ++n)
{
for(int k = 0; k < K_pk; ++k)
{
int n0 = n / NLane;
int n1 = n % NLane;
int k0 = k / (KLane * KPack);
tempk = k % (KLane * KPack);
int k1 = tempk / KPack;
int k2 = tempk % KPack;
int outputIndex = n0 * KPack * NLane * KLane * K0 + k0 * KPack * NLane * KLane +
k1 * KPack * NLane + n1 * KPack + k2;
dst[outputIndex] = src[n * K_pk + k];
}
}
}
float i4_to_f32_gfx9(uint8_t i4)
{
static std::unordered_map<uint8_t, float> u = {{0b1000, -0.5000f},
{0b1001, -0.4375f},
{0b1010, -0.3750f},
{0b1011, -0.3125f},
{0b1100, -0.2500f},
{0b1101, -0.1875f},
{0b1110, -0.1250f},
{0b1111, -0.0625f},
{0b0, +0.0000f},
{0b1, +0.0625f},
{0b10, +0.1250f},
{0b11, +0.1875f},
{0b100, +0.2500f},
{0b101, +0.3125f},
{0b110, +0.3750f},
{0b111, +0.4375f}};
return u[i4];
}
inline void permute_b_pk_i4(Tensor<ck::pk_i4_t>& b_k_n_permute,
int N,
int K,
Tensor<float>& b_k_n_f32,
Tensor<float>& b_k_n_gfx9_f32)
{
for(int n = 0; n < N; n++)
{
for(int k = 0; k < K; k++)
{
ck::pk_i4_t i4x2 = b_k_n_permute(k, n).data;
uint8_t i4 = 0;
if(k % 2 == 1)
i4 = (i4x2.data >> 0) & 0xf;
else
i4 = (i4x2.data >> 4) & 0xf;
b_k_n_f32(k, n) = (((i4 & 0x0f) >> 0) - 8.f);
b_k_n_gfx9_f32(k, n) = i4_to_f32_gfx9(i4);
}
}
// vector pk_i4x4 permute
for(int i = 0; i < N; i++)
{
for(int j = 0; j < K; j += 8)
{
int input[8];
for(int k = 0; k < 4; k++)
{
int i4x2 = b_k_n_permute(j + k * 2, i).data;
input[k * 2 + 0] = (i4x2 >> 4) & 0xf;
input[k * 2 + 1] = (i4x2 >> 0) & 0xf;
}
// permute 01234567->20643175
{
int hi = input[2];
int lo = input[0];
int i4x2 = (hi << 4) | lo;
b_k_n_permute(j + 0, i) = i4x2;
}
{
int hi = input[6];
int lo = input[4];
int i4x2 = (hi << 4) | lo;
b_k_n_permute(j + 2, i) = i4x2;
}
{
int hi = input[3];
int lo = input[1];
int i4x2 = (hi << 4) | lo;
b_k_n_permute(j + 4, i) = i4x2;
}
{
int hi = input[7];
int lo = input[5];
int i4x2 = (hi << 4) | lo;
b_k_n_permute(j + 6, i) = i4x2;
}
}
}
}
inline void permute_a_pk_i4(Tensor<ck::pk_i4_t>& a_m_k_permute,
int M,
int K,
Tensor<float>& a_m_k_f32,
Tensor<float>& a_m_k_gfx9_f32)
{
for(int m = 0; m < M; m++)
{
for(int k = 0; k < K; k++)
{
ck::pk_i4_t i4x2 = a_m_k_permute(m, k).data;
uint8_t i4 = 0;
if(k % 2 == 1)
i4 = (i4x2.data >> 0) & 0xf;
else
i4 = (i4x2.data >> 4) & 0xf;
a_m_k_f32(m, k) = (((i4 & 0x0f) >> 0) - 8.f);
a_m_k_gfx9_f32(m, k) = i4_to_f32_gfx9(i4);
}
}
// vector pk_i4x4 permute
for(int i = 0; i < M; i++)
{
for(int j = 0; j < K; j += 8)
{
int input[8];
for(int k = 0; k < 4; k++)
{
int i4x2 = a_m_k_permute(i, j + k * 2).data;
input[k * 2 + 0] = (i4x2 >> 4) & 0xf;
input[k * 2 + 1] = (i4x2 >> 0) & 0xf;
}
// permute 01234567->20643175
{
int hi = input[2];
int lo = input[0];
int i4x2 = (hi << 4) | lo;
a_m_k_permute(i, j + 0) = i4x2;
}
{
int hi = input[6];
int lo = input[4];
int i4x2 = (hi << 4) | lo;
a_m_k_permute(i, j + 2) = i4x2;
}
{
int hi = input[3];
int lo = input[1];
int i4x2 = (hi << 4) | lo;
a_m_k_permute(i, j + 4) = i4x2;
}
{
int hi = input[7];
int lo = input[5];
int i4x2 = (hi << 4) | lo;
a_m_k_permute(i, j + 6) = i4x2;
}
}
}
}
Reference in New Issue View Git Blame Copy Permalink