diff --git a/include/ck/tensor_operation/gpu/device/impl/device_conv2d_backward_weight_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp b/include/ck/tensor_operation/gpu/device/impl/device_conv2d_backward_weight_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
index 1e50e1ae89..23440e24f6 100644
--- a/include/ck/tensor_operation/gpu/device/impl/device_conv2d_backward_weight_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
+++ b/include/ck/tensor_operation/gpu/device/impl/device_conv2d_backward_weight_xdl_c_shuffle_nhwc_kyxc_nhwk.hpp
@@ -532,11 +532,12 @@ struct DeviceConv2dBwdWeightXdl_C_Shuffle_Input_N_Hi_Wi_C_Weight_K_Y_X_C_Output_
             float ave_time = 0;
 
             const auto Run = [&](const auto& kernel) {
-                hipGetErrorString(hipMemset(
+                hipGetErrorString(hipMemsetAsync(
                     arg.p_c_grid_,
                     0,
                     arg.c_grid_desc_mblock_mperblock_nblock_nperblock_.GetElementSpaceSize() *
-                        sizeof(CDataType)));
+                        sizeof(CDataType),
+                    stream_config.stream_id_));
 
                 ave_time =
                     launch_and_time_kernel(stream_config,
diff --git a/include/ck/tensor_operation/gpu/device/impl/device_gemm_xdl_splitk_c_shuffle.hpp b/include/ck/tensor_operation/gpu/device/impl/device_gemm_xdl_splitk_c_shuffle.hpp
index 2b8f760737..505e1846d1 100644
--- a/include/ck/tensor_operation/gpu/device/impl/device_gemm_xdl_splitk_c_shuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/impl/device_gemm_xdl_splitk_c_shuffle.hpp
@@ -158,8 +158,10 @@ struct DeviceGemmXdlSplitKCShuffle : public DeviceGemmSplitK<ALayout,
 
             const auto Run = [&](const auto& kernel) {
                 if(kbatch > 1)
-                    hipGetErrorString(
-                        hipMemset(karg.p_c_grid, 0, karg.M * karg.N * sizeof(CDataType)));
+                    hipGetErrorString(hipMemsetAsync(karg.p_c_grid,
+                                                     0,
+                                                     karg.M * karg.N * sizeof(CDataType),
+                                                     stream_config.stream_id_));
 
                 ave_time = launch_and_time_kernel(
                     stream_config, kernel, dim3(gdx, gdy, gdz), dim3(BlockSize), 0, karg, b2c_map);
diff --git a/include/ck/tensor_operation/gpu/device/impl/device_gemm_xdl_streamk.hpp b/include/ck/tensor_operation/gpu/device/impl/device_gemm_xdl_streamk.hpp
index 6fe5784c28..8de42ba9ef 100644
--- a/include/ck/tensor_operation/gpu/device/impl/device_gemm_xdl_streamk.hpp
+++ b/include/ck/tensor_operation/gpu/device/impl/device_gemm_xdl_streamk.hpp
@@ -1,5 +1,5 @@
 // SPDX-License-Identifier: MIT
-// Copyright (c) 2018-2022, Advanced Micro Devices, Inc. All rights reserved.
+// Copyright (c) 2018-2023, Advanced Micro Devices, Inc. All rights reserved.
 
 #pragma once
 
@@ -147,7 +147,10 @@ struct DeviceGemmXdlStreamK : public DeviceGemmStreamK<ALayout,
             if constexpr(GridwiseGemm::Block2CTileMap::ReductionStrategy ==
                          StreamKReductionStrategy::Atomic)
             {
-                hipGetErrorString(hipMemset(karg.p_c_grid, 0, karg.M * karg.N * sizeof(CDataType)));
+                hipGetErrorString(hipMemsetAsync(karg.p_c_grid,
+                                                 0,
+                                                 karg.M * karg.N * sizeof(CDataType),
+                                                 stream_config.stream_id_));
                 ave_time = launch_and_time_kernel(stream_config,
                                                   kernel,
                                                   grid_dims,
diff --git a/include/ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl_splitk_cshuffle.hpp b/include/ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl_splitk_cshuffle.hpp
index 4f4413b78f..ba66d8358c 100644
--- a/include/ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl_splitk_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/impl/device_grouped_gemm_xdl_splitk_cshuffle.hpp
@@ -421,8 +421,10 @@ struct DeviceGroupedGemmXdlSplitKCShuffle : public DeviceGroupedGemmSplitK<ALayo
                     for(const auto& trans_arg : arg.gemm_kernel_args_)
                     {
                         const auto& karg = trans_arg.karg_;
-                        hip_check_error(
-                            hipMemset(karg.p_c_grid, 0, karg.M * karg.N * sizeof(EDataType)));
+                        hip_check_error(hipMemsetAsync(karg.p_c_grid,
+                                                       0,
+                                                       karg.M * karg.N * sizeof(EDataType),
+                                                       stream_config.stream_id_));
                     }
                 }
 
diff --git a/include/ck/tensor_operation/gpu/device/impl/device_splitk_contraction_multiple_d_xdl_cshuffle.hpp b/include/ck/tensor_operation/gpu/device/impl/device_splitk_contraction_multiple_d_xdl_cshuffle.hpp
index 7ba2d96a9c..a52968bcef 100644
--- a/include/ck/tensor_operation/gpu/device/impl/device_splitk_contraction_multiple_d_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/impl/device_splitk_contraction_multiple_d_xdl_cshuffle.hpp
@@ -886,11 +886,12 @@ struct DeviceSplitKContractionMultipleD_Xdl_CShuffle
                     typename GridwiseGemmAtomicAdd::DefaultBlock2ETileMap,
                     has_main_loop>;
 
-                hipGetErrorString(hipMemset(
+                hipGetErrorString(hipMemsetAsync(
                     arg.p_e_grid_,
                     0,
                     arg.e_grid_desc_mblock_mperblock_nblock_nperblock_.GetElementSpaceSize() *
-                        sizeof(EDataType)));
+                        sizeof(EDataType),
+                    stream_config.stream_id_));
 
                 return launch_and_time_kernel(stream_config,
                                               kernel,