Implement DPP8 based GEMM for Navi21 (#826)

ROCm · Aug 14, 2023 · d4c8425 · d4c8425
1 parent f60f0a5
commit d4c8425
Show file tree

Hide file tree

Showing 18 changed files with 1,217 additions and 30 deletions.
diff --git a/example/01_gemm/CMakeLists.txt b/example/01_gemm/CMakeLists.txt
@@ -6,6 +6,8 @@ if(DL_KERNELS)
   if(DTYPES MATCHES "fp16" OR NOT DEFINED DTYPES)
     add_example_executable(example_gemm_dl_fp16 gemm_dl_fp16.cpp)
     add_dependencies(example_gemm_dl example_gemm_dl_fp16)
+    add_example_executable(example_gemm_dl_dpp8_fp16 gemm_dl_dpp8_fp16.cpp)
+    add_dependencies(example_gemm_dl example_gemm_dl_dpp8_fp16)
   endif()
   if(DTYPES MATCHES "int8" OR NOT DEFINED DTYPES)
     add_example_executable(example_gemm_dl_int8 gemm_dl_int8.cpp)

diff --git a/example/01_gemm/gemm_dl_dpp8_fp16.cpp b/example/01_gemm/gemm_dl_dpp8_fp16.cpp
@@ -0,0 +1,37 @@
+// SPDX-License-Identifier: MIT
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "common.hpp"
+
+#include "ck/tensor_operation/gpu/device/impl/device_gemm_dl_dpp8.hpp"
+
+using ADataType   = ck::half_t;
+using BDataType   = ck::half_t;
+using CDataType   = ck::half_t;
+using AccDataType = float;
+
+using ALayout = Col;
+using BLayout = Row;
+using CLayout = Row;
+
+using AElementOp = PassThrough;
+using BElementOp = PassThrough;
+using CElementOp = PassThrough;
+
+static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
+
+// clang-format off
+using DeviceGemmInstance = ck::tensor_operation::device::DeviceGemmDlDpp8
+// ######|     AData|     BData|     CData|     AccData| ALayout| BLayout| CLayout|           A|           B|           C|           GEMM| Block|  MPer|  NPer| K0Per| K1|      M1Per|      N1Per|   KPer|  M11N11Thread|  M11N11Thread|     ABlockTransfer|       ABlockTransfer| ABlockTransfer| ABlockTransfer|      ABlockTransfer|     ABlockTransfer|      ABlockTransfer|     BBlockTransfer|       BBlockTransfer| BBlockTransfer| BBlockTransfer|      BBlockTransfer|     BBlockTransfer|      BBlockTransfer|     CThreadTransfer| CThreadTransfer|    CThreadTransfer|
+// ######|      Type|      Type|      Type|        Type|        |        |        | Elementwise| Elementwise| Elementwise| Spacialization|  Size| Block| Block| Block|   | ThreadM111| ThreadN111| Thread| ClusterM110Xs| ClusterN110Xs| ThreadSliceLengths| ThreadClusterLengths|  ThreadCluster|      SrcAccess|     SrcVectorTensor|    SrcVectorTensor|     DstVectorTensor| ThreadSliceLengths| ThreadClusterLengths|  ThreadCluster|      SrcAccess|     SrcVectorTensor|    SrcVectorTensor|     DstVectorTensor|        SrcDstAccess| SrcDstVectorDim| DstScalarPerVector|
+// ######|          |          |          |            |        |        |        |   Operation|   Operation|   Operation|               |      |      |      |      |   |           |           |       |              |              |        K0_M0_M1_K1|          K0_M0_M1_K1|   ArrangeOrder|          Order| Lengths_K0_M0_M1_K1| ContiguousDimOrder| Lengths_K0_M0_M1_K1|        K0_N0_N1_K1|          K0_N0_N1_K1|   ArrangeOrder|          Order| Lengths_K0_N0_N1_K1| ContiguousDimOrder| Lengths_K0_N0_N1_K1|               Order|                |                   |
+// ######|          |          |          |            |        |        |        |            |            |            |               |      |      |      |      |   |           |           |       |              |              |                   |                     |               |               |                    |                   |                    |                   |                     |               |               |                    |                   |                    |                    |                |                   |
+         < ADataType, BDataType, CDataType, AccDataType, ALayout, BLayout, CLayout,  AElementOp,  BElementOp,  CElementOp,    GemmDefault,   256,   128,   128,    16,  2,          1,          8,      8,       S<8, 8>,       S<4, 1>,      S<2, 1, 4, 2>,      S<8, 1,  32, 1>,  S<0, 3, 1, 2>,  S<0, 3, 1, 2>,       S<1, 1, 4, 1>,      S<0, 3, 1, 2>,       S<1, 1, 4, 2>,      S<2, 1, 4, 2>,       S<8, 1, 32, 1>,  S<0, 3, 1, 2>,  S<0, 3, 1, 2>,       S<1, 1, 4, 1>,      S<0, 3, 1, 2>,       S<1, 1, 4, 2>, S<0, 1, 2, 3, 4, 5>,               5,                  4>;
+// clang-format on
+
+using ReferenceGemmInstance = ck::tensor_operation::host::
+    ReferenceGemm<ADataType, BDataType, CDataType, AccDataType, AElementOp, BElementOp, CElementOp>;
+
+#include "run_gemm_example.inc"
+
+int main(int argc, char* argv[]) { return !run_gemm_example(argc, argv); }
diff --git a/include/ck/ck.hpp b/include/ck/ck.hpp
@@ -125,6 +125,9 @@
 // `s_nop`s to avoid hazard
 #define CK_USE_AMD_V_DOT_INLINE_ASM 0
 
+// inner product using V_DOT with DPP8 modifiers
+#define CK_USE_AMD_V_DOT_DPP8_INLINE_ASM 1
+
 // block synchronization only s_wait lgkmcnt(0), not vmcnt(0)
 #define CK_EXPERIMENTAL_BLOCK_SYNC_LDS_WITHOUT_SYNC_VMEM 1