added gemm xdl fp16_int8_fp16

example/01_gemm/CMakeLists.txt

@@ -71,3 +71,6 @@ endforeach()
 
 add_example_executable(example_gemm_xdl_fp16_fp8 gemm_xdl_fp16_fp8.cpp)
 add_example_dependencies(example_gemm_xdl example_gemm_xdl_fp16_fp8)
+
+add_example_executable(example_gemm_xdl_fp16_int8 gemm_xdl_fp16_int8.cpp)
+add_example_dependencies(example_gemm_xdl example_gemm_xdl_fp16_int8)

example/01_gemm/gemm_xdl_fp16_int8.cpp

+// SPDX-License-Identifier: MIT
+// Copyright (c) 2024, Advanced Micro Devices, Inc. All rights reserved.
+
+#include "common.hpp"
+
+#include "ck/tensor_operation/gpu/device/impl/device_gemm_xdl_cshuffle.hpp"
+
+using ADataType        = ck::half_t;
+using BDataType        = int8_t;
+using CDataType        = ck::half_t;
+using AccDataType      = float;
+using CShuffleDataType = ck::half_t;
+
+using ALayout = Row;
+using BLayout = Col;
+using CLayout = Row;
+
+using AElementOp = PassThrough;
+using BElementOp = PassThrough;
+using CElementOp = PassThrough;
+
+static constexpr auto GemmDefault = ck::tensor_operation::device::GemmSpecialization::Default;
+static constexpr auto LoopSched   = ck::make_default_loop_scheduler();
+static constexpr auto PipelineVer = ck::PipelineVersion::v1;
+using ComputeType                 = ck::half_t;
+
+// clang-format off
+using DeviceGemmInstance = ck::tensor_operation::device::DeviceGemm_Xdl_CShuffle
+// ######| ALayout| BLayout| CLayout|     AData|     BData|     CData|     AccData|         CShuffle|           A|           B|           C|           GEMM| NumGemmK| Block|  MPer|  NPer|  KPer| AK1| BK1| MPer| NPer| MXdl| NXdl|  ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockTransfer| ABlockLds|  BBlockTransfer| BBlockTransfer| BBlockTransfer| BlockTransfer| BBlockTransfer| BBlockTransfer| BBlockLds|    CShuffle|    CShuffle| CBlockTransferClusterLengths|  CBlockTransfer|      Loop|    Pipeline|  ComputeType|
+// ######|        |        |        |      Type|      Type|      Type|        Type|         DataType| Elementwise| Elementwise| Elementwise| Spacialization| Prefetch|  Size| Block| Block| Block|    |    |  XDL|  XDL|  Per|  Per|   ThreadCluster|  ThreadCluster| SrcAccessOrder|   SrcVectorDim|      SrcScalar|      DstScalar| AddExtraM|   ThreadCluster|  ThreadCluster| SrcAccessOrder|  SrcVectorDim|      SrcScalar|      DstScalar| AddExtraN| MXdlPerWave| NXdlPerWave|         _MBlock_MWaveMPerXdl| ScalarPerVector| Scheduler|     Version|             |
+// ######|        |        |        |          |          |          |            |                 |   Operation|   Operation|   Operation|               |    Stage|      |      |      |      |    |    |     |     | Wave| Wave| Lengths_K0_M_K1|   ArrangeOrder|               |               |      PerVector|   PerVector_K1|          | Lengths_K0_N_K1|   ArrangeOrder|               |              |      PerVector|   PerVector_K1|          |  PerShuffle|  PerShuffle|         _NBlock_NWaveNPerXdl|   _NWaveNPerXdl|          |            |             |
+// ######|        |        |        |          |          |          |            |                 |            |            |            |               |         |      |      |      |      |    |    |     |     |     |     |                |               |               |               |               |               |          |                |               |               |              |               |               |          |            |            |                             |                |          |            |             |
+        < ALayout, BLayout, CLayout, ADataType, BDataType, CDataType, AccDataType, CShuffleDataType,  AElementOp,  BElementOp,  CElementOp,    GemmDefault,        1,   256,   256,   128,    32,   8,   8,   32,   32,    4,    2,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,              2,              8,              8,         1,     S<4, 64, 1>,     S<1, 0, 2>,     S<1, 0, 2>,             2,              8,              8,         1,           1,           1,               S<1, 32, 1, 8>,               8,  LoopSched, PipelineVer, ComputeType>;
+// 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); }

library/include/ck/library/tensor_operation_instance/gpu/gemm.hpp

@@ -367,6 +367,17 @@ void add_device_gemm_xdl_c_shuffle_f16_f8_f16_mk_nk_mn_instances(
         DeviceGemm<Row, Col, Row, F16, F8, F16, PassThrough, PassThrough, PassThrough>>>&
         instances);
 #endif
+#if defined(CK_ENABLE_FP16) && defined(CK_ENABLE_INT8)
+void add_device_gemm_xdl_c_shuffle_f16_int8_f16_mk_kn_mn_instances(
+    std::vector<std::unique_ptr<
+        DeviceGemm<Row, Row, Row, F16, I8, F16, PassThrough, PassThrough, PassThrough>>>&
+        instances);
+
+void add_device_gemm_xdl_c_shuffle_f16_int8_f16_mk_nk_mn_instances(
+    std::vector<std::unique_ptr<
+        DeviceGemm<Row, Row, Row, F16, I8, F16, PassThrough, PassThrough, PassThrough>>>&
+        instances);
+#endif
 
 template <typename ALayout,
           typename BLayout,
@@ -612,6 +623,21 @@ struct DeviceOperationInstanceFactory<
                 add_device_gemm_xdl_c_shuffle_f16_f8_f16_mk_nk_mn_instances(op_ptrs);
             }
         }
+#endif
+#if defined(CK_ENABLE_FP16) && defined(CK_ENABLE_INT8)
+        else if constexpr(is_same_v<ADataType, half_t> && is_same_v<BDataType, int8_t> &&
+                          is_same_v<CDataType, half_t>)
+        {
+            if constexpr(is_same_v<ALayout, Row> && is_same_v<BLayout, Row> &&
+                         is_same_v<CLayout, Row>)
+            {
+                add_device_gemm_xdl_c_shuffle_f16_int8_f16_mk_kn_mn_instances(op_ptrs);
+            }
+            else if constexpr(is_same_v<ALayout, Row> && is_same_v<BLayout, Col> &&
+                         is_same_v<CLayout, Row>)
+            {
+                add_device_gemm_xdl_c_shuffle_f16_int8_f16_mk_nk_mn_instances(op_ptrs);
+            }
 #endif
         return op_ptrs;
     }

library/src/tensor_operation_instance/gpu/gemm/CMakeLists.txt

@@ -46,6 +46,8 @@ list(APPEND GEMM_INSTANCES
     device_gemm_xdl_c_shuffle_f16_f16_f16_km_nk_mn_instance.cpp
     device_gemm_xdl_c_shuffle_2_stage_f16_f16_f16_mk_nk_mn_instance.cpp
     device_gemm_xdl_c_shuffle_lds_direct_load_f16_f16_f16_mk_nk_mn_instance.cpp
+    device_gemm_xdl_c_shuffle_f16_int8_f16_mk_kn_mn_instance.cpp
+    device_gemm_xdl_c_shuffle_f16_int8_f16_mk_nk_mn_instance.cpp
     device_gemm_xdl_f16_f16_f16/km_kn_mn_add_instance.cpp
     device_gemm_xdl_f16_f16_f16/km_kn_mn_default_pipeline_v1_instance.cpp
     device_gemm_xdl_f16_f16_f16/km_kn_mn_default_pipeline_v2_instance.cpp

profiler/src/profile_gemm.cpp

@@ -24,6 +24,7 @@ enum struct GemmDataType
     BF16_BF16_BF16, // 2
     INT8_INT8_INT8, // 3
     F8_F8_F8,       // 4
+    F16_INT8_F16    // 5
 };
 
 #define OP_NAME "gemm"
@@ -32,7 +33,7 @@ enum struct GemmDataType
 static void print_helper_msg()
 {
     std::cout << "arg1: tensor operation (" OP_NAME ": " OP_DESC ")\n"
-              << "arg2: data type (0: fp32; 1: fp16; 2: bf16; 3: int8; 4: fp8)\n"
+              << "arg2: data type (0: fp32; 1: fp16; 2: bf16; 3: int8; 4: fp8; 5: fp16 & int8)\n"
               << "arg3: matrix layout (0: A[m, k] * B[k, n] = C[m, n];\n"
               << "                     1: A[m, k] * B[n, k] = C[m, n];\n"
               << "                     2: A[k, m] * B[k, n] = C[m, n];\n"
@@ -175,6 +176,16 @@ int profile_gemm(int argc, char* argv[])
         return profile(Col{}, Col{}, Row{}, F16{}, F16{}, F32{}, F16{});
     }
 #endif
+#if defined(CK_ENABLE_FP16) && defined(CK_ENABLE_INT8)
+    else if(data_type == GemmDataType::F16_INT8_F16 && layout == GemmMatrixLayout::MK_KN_MN)
+    {
+        return profile(Row{}, Row{}, Row{}, F16{}, INT8{}, F32{}, F16{});
+    }
+    else if(data_type == GemmDataType::F16_INT8_F16 && layout == GemmMatrixLayout::MK_NK_MN)
+    {
+        return profile(Row{}, Col{}, Row{}, F16{}, INT8{}, F32{}, F16{});
+    }
+#endif
 #ifdef CK_ENABLE_BF16
     else if(data_type == GemmDataType::BF16_BF16_BF16 && layout == GemmMatrixLayout::MK_KN_MN)
     {