diff --git a/example/01_gemm/CMakeLists.txt b/example/01_gemm/CMakeLists.txt
index 5b71cd1548..042115197a 100644
--- a/example/01_gemm/CMakeLists.txt
+++ b/example/01_gemm/CMakeLists.txt
@@ -59,7 +59,7 @@ add_example_dependencies(example_gemm_xdl example_gemm_xdl_fp8)
 add_example_executable(example_gemm_xdl_fp8_bf8 gemm_xdl_fp8_bf8.cpp)
 add_example_dependencies(example_gemm_xdl example_gemm_xdl_fp8_bf8)
 
-list(APPEND gpu_list gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
     if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/02_gemm_bilinear/CMakeLists.txt b/example/02_gemm_bilinear/CMakeLists.txt
index d82c42d5a9..2da534f278 100644
--- a/example/02_gemm_bilinear/CMakeLists.txt
+++ b/example/02_gemm_bilinear/CMakeLists.txt
@@ -1,5 +1,5 @@
 list(APPEND gpu_list1 gfx1100 gfx1101 gfx1102)
-list(APPEND gpu_list2 gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list2 gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list1 AND target EQUAL 0)
diff --git a/example/03_gemm_bias_relu/CMakeLists.txt b/example/03_gemm_bias_relu/CMakeLists.txt
index 2f5cba924d..73bdfce535 100644
--- a/example/03_gemm_bias_relu/CMakeLists.txt
+++ b/example/03_gemm_bias_relu/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/04_gemm_add_add_fastgelu/CMakeLists.txt b/example/04_gemm_add_add_fastgelu/CMakeLists.txt
index 33ac1e7e77..ddc0916074 100644
--- a/example/04_gemm_add_add_fastgelu/CMakeLists.txt
+++ b/example/04_gemm_add_add_fastgelu/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
     if(gpu IN_LIST gpu_list AND target EQUAL 0)
@@ -33,4 +33,4 @@ foreach(gpu IN LISTS GPU_TARGETS)
         add_example_dependencies(example_gemm_add_add_fastgelu_xdl example_gemm_add_add_fastgelu_xdl_lds_direct_load_fp32)
         set(target 1)
     endif()
-endforeach()
\ No newline at end of file
+endforeach()
diff --git a/example/09_convnd_fwd/CMakeLists.txt b/example/09_convnd_fwd/CMakeLists.txt
index f9903bfe03..502ba59bee 100644
--- a/example/09_convnd_fwd/CMakeLists.txt
+++ b/example/09_convnd_fwd/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/10_convnd_fwd_multiple_d_multiple_reduce/CMakeLists.txt b/example/10_convnd_fwd_multiple_d_multiple_reduce/CMakeLists.txt
index 222a3b7c0b..4002f7ca52 100644
--- a/example/10_convnd_fwd_multiple_d_multiple_reduce/CMakeLists.txt
+++ b/example/10_convnd_fwd_multiple_d_multiple_reduce/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
     if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/14_gemm_quantization/CMakeLists.txt b/example/14_gemm_quantization/CMakeLists.txt
index 9793e8b8a0..b38b11be0f 100644
--- a/example/14_gemm_quantization/CMakeLists.txt
+++ b/example/14_gemm_quantization/CMakeLists.txt
@@ -1,7 +1,7 @@
 # dlops
 add_example_executable(example_gemm_dl_quantization_int8 gemm_dl_quantization_int8.cpp)
 # xdlops
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/16_gemm_multi_d_multi_reduces/CMakeLists.txt b/example/16_gemm_multi_d_multi_reduces/CMakeLists.txt
index 5955e1d6cb..80c1022ecd 100644
--- a/example/16_gemm_multi_d_multi_reduces/CMakeLists.txt
+++ b/example/16_gemm_multi_d_multi_reduces/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
     if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/17_convnd_bwd_data/CMakeLists.txt b/example/17_convnd_bwd_data/CMakeLists.txt
index 7c6d10d8a0..155d9ad77f 100644
--- a/example/17_convnd_bwd_data/CMakeLists.txt
+++ b/example/17_convnd_bwd_data/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/18_batched_gemm_reduce/CMakeLists.txt b/example/18_batched_gemm_reduce/CMakeLists.txt
index 94ed129dc0..018b57f82c 100644
--- a/example/18_batched_gemm_reduce/CMakeLists.txt
+++ b/example/18_batched_gemm_reduce/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/20_grouped_conv_bwd_weight/CMakeLists.txt b/example/20_grouped_conv_bwd_weight/CMakeLists.txt
index c28fca6fa2..a418bafefb 100644
--- a/example/20_grouped_conv_bwd_weight/CMakeLists.txt
+++ b/example/20_grouped_conv_bwd_weight/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list_xdl gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list_xdl gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 list(APPEND gpu_list_wmma gfx1100 gfx1101 gfx1102)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
diff --git a/example/21_gemm_layernorm/CMakeLists.txt b/example/21_gemm_layernorm/CMakeLists.txt
index e231bc619b..57e7eefd7c 100644
--- a/example/21_gemm_layernorm/CMakeLists.txt
+++ b/example/21_gemm_layernorm/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/30_grouped_conv_fwd_multiple_d/CMakeLists.txt b/example/30_grouped_conv_fwd_multiple_d/CMakeLists.txt
index 3a8c2ef52f..209336b2de 100644
--- a/example/30_grouped_conv_fwd_multiple_d/CMakeLists.txt
+++ b/example/30_grouped_conv_fwd_multiple_d/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list1 gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list1 gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 list(APPEND gpu_list2 gfx1100 gfx1101 gfx1102)
 
 set(target 0)
diff --git a/example/31_batched_gemm_gemm/CMakeLists.txt b/example/31_batched_gemm_gemm/CMakeLists.txt
index 93f16c945f..149013064e 100644
--- a/example/31_batched_gemm_gemm/CMakeLists.txt
+++ b/example/31_batched_gemm_gemm/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list1 gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list1 gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
@@ -13,6 +13,6 @@ foreach(gpu IN LISTS GPU_TARGETS)
  endif()
 endforeach()
 
-if(NOT GPU_TARGETS MATCHES "gfx94" AND NOT GPU_TARGETS MATCHES "gfx1")
+if(NOT GPU_TARGETS MATCHES "gfx94" AND NOT GPU_TARGETS MATCHES "gfx95" AND NOT GPU_TARGETS MATCHES "gfx1")
    add_example_executable(example_batched_gemm_gemm_xdl_int8 batched_gemm_gemm_xdl_int8.cpp)
 endif()
diff --git a/example/35_splitK_gemm/CMakeLists.txt b/example/35_splitK_gemm/CMakeLists.txt
index f98308d687..f724d4e9b4 100644
--- a/example/35_splitK_gemm/CMakeLists.txt
+++ b/example/35_splitK_gemm/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/38_grouped_conv_bwd_data_multiple_d/CMakeLists.txt b/example/38_grouped_conv_bwd_data_multiple_d/CMakeLists.txt
index 1ae179e950..99b76cb9cb 100644
--- a/example/38_grouped_conv_bwd_data_multiple_d/CMakeLists.txt
+++ b/example/38_grouped_conv_bwd_data_multiple_d/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list_xdl gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list_xdl gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 list(APPEND gpu_list_wmma gfx1100 gfx1101 gfx1102)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
diff --git a/example/40_conv2d_fwd_quantization/CMakeLists.txt b/example/40_conv2d_fwd_quantization/CMakeLists.txt
index 2d804cafe9..48a3f052bc 100644
--- a/example/40_conv2d_fwd_quantization/CMakeLists.txt
+++ b/example/40_conv2d_fwd_quantization/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/41_grouped_conv_conv_fwd/CMakeLists.txt b/example/41_grouped_conv_conv_fwd/CMakeLists.txt
index ae251e88d2..f8140a19d4 100644
--- a/example/41_grouped_conv_conv_fwd/CMakeLists.txt
+++ b/example/41_grouped_conv_conv_fwd/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list1 gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list1 gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 list(APPEND gpu_list2 gfx908 gfx90a)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
@@ -13,6 +13,6 @@ foreach(gpu IN LISTS GPU_TARGETS)
  endif()
 endforeach()
 
-if(NOT GPU_TARGETS MATCHES "gfx94" AND NOT GPU_TARGETS MATCHES "gfx1")
+if(NOT GPU_TARGETS MATCHES "gfx94" AND NOT GPU_TARGETS MATCHES "gfx95" AND NOT GPU_TARGETS MATCHES "gfx1")
    add_example_executable(example_grouped_conv_conv_fwd_xdl_int8 grouped_conv_conv_fwd_xdl_int8.cpp)
 endif()
diff --git a/example/44_elementwise_permute/CMakeLists.txt b/example/44_elementwise_permute/CMakeLists.txt
index a963399dc7..bd100fa650 100644
--- a/example/44_elementwise_permute/CMakeLists.txt
+++ b/example/44_elementwise_permute/CMakeLists.txt
@@ -5,6 +5,6 @@ add_example_executable(example_elementwise_permute_4D_fp16_row elementwise_permu
 add_example_executable(example_elementwise_permute_4D_fp32_col elementwise_permute_4D_fp32_col.cpp)
 add_example_executable(example_elementwise_permute_4D_fp16_col elementwise_permute_4D_fp16_col.cpp)
 add_example_executable(example_elementwise_permute elementwise_permute.cpp)
-if((NOT GPU_TARGETS MATCHES "gfx940") AND (NOT GPU_TARGETS MATCHES "gfx941") AND (NOT GPU_TARGETS MATCHES "gfx942"))
+if((NOT GPU_TARGETS MATCHES "gfx940") AND (NOT GPU_TARGETS MATCHES "gfx941") AND (NOT GPU_TARGETS MATCHES "gfx942") AND (NOT GPU_TARGETS MATCHES "gfx950"))
     add_example_executable(example_elementwise_permute_3d elementwise_permute_3d.cpp)
 endif()
diff --git a/example/47_gemm_bias_softmax_gemm_permute/CMakeLists.txt b/example/47_gemm_bias_softmax_gemm_permute/CMakeLists.txt
index 14432f6e23..67534b291a 100644
--- a/example/47_gemm_bias_softmax_gemm_permute/CMakeLists.txt
+++ b/example/47_gemm_bias_softmax_gemm_permute/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/52_im2col_col2im/CMakeLists.txt b/example/52_im2col_col2im/CMakeLists.txt
index 4dc6c8b4e0..4396207cdb 100644
--- a/example/52_im2col_col2im/CMakeLists.txt
+++ b/example/52_im2col_col2im/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
     if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/example/60_gemm_multi_ABD/CMakeLists.txt b/example/60_gemm_multi_ABD/CMakeLists.txt
index 57bc0b33ef..610e8bc876 100644
--- a/example/60_gemm_multi_ABD/CMakeLists.txt
+++ b/example/60_gemm_multi_ABD/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list2 gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list2 gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
     if(gpu IN_LIST gpu_list2 AND target EQUAL 0)
diff --git a/example/61_contraction_multi_ABD/CMakeLists.txt b/example/61_contraction_multi_ABD/CMakeLists.txt
index 42500b64e6..a6094fbe40 100644
--- a/example/61_contraction_multi_ABD/CMakeLists.txt
+++ b/example/61_contraction_multi_ABD/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list2 gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list2 gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list2 AND target EQUAL 0)
diff --git a/example/62_conv_fwd_activ/CMakeLists.txt b/example/62_conv_fwd_activ/CMakeLists.txt
index d1f26bbfe1..15d08846d5 100644
--- a/example/62_conv_fwd_activ/CMakeLists.txt
+++ b/example/62_conv_fwd_activ/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/include/ck/ck.hpp b/include/ck/ck.hpp
index c93d1d0639..628da29815 100644
--- a/include/ck/ck.hpp
+++ b/include/ck/ck.hpp
@@ -45,7 +45,7 @@
 #endif
 
 // define general macros for various architectures
-#if defined(__gfx940__) || defined(__gfx941__) || defined(__gfx942__)
+#if defined(__gfx940__) || defined(__gfx941__) || defined(__gfx942__) || defined(__gfx950__)
 #define __gfx94__
 #endif
 #if defined(__gfx1010__) || defined(__gfx1011__) || defined(__gfx1012__)
diff --git a/include/ck/host_utility/device_prop.hpp b/include/ck/host_utility/device_prop.hpp
index 13e5268752..a0aa68b608 100644
--- a/include/ck/host_utility/device_prop.hpp
+++ b/include/ck/host_utility/device_prop.hpp
@@ -55,14 +55,15 @@ inline bool is_xdl_supported()
 {
     return ck::get_device_name() == "gfx908" || ck::get_device_name() == "gfx90a" ||
            ck::get_device_name() == "gfx940" || ck::get_device_name() == "gfx941" ||
-           ck::get_device_name() == "gfx942";
+           ck::get_device_name() == "gfx942" || ck::get_device_name() == "gfx950";
 }
 
 inline bool is_lds_direct_load_supported()
 {
     // Check if direct loads from global memory to LDS are supported.
     return ck::get_device_name() == "gfx90a" || ck::get_device_name() == "gfx940" ||
-           ck::get_device_name() == "gfx941" || ck::get_device_name() == "gfx942";
+           ck::get_device_name() == "gfx941" || ck::get_device_name() == "gfx942" ||
+           ck::get_device_name() == "gfx950";
 }
 
 inline bool is_navi1_supported()
diff --git a/include/ck/tensor_operation/gpu/device/impl/device_contraction_multiple_d_xdl_cshuffle.hpp b/include/ck/tensor_operation/gpu/device/impl/device_contraction_multiple_d_xdl_cshuffle.hpp
index 1f65afed3d..8a84d031e7 100644
--- a/include/ck/tensor_operation/gpu/device/impl/device_contraction_multiple_d_xdl_cshuffle.hpp
+++ b/include/ck/tensor_operation/gpu/device/impl/device_contraction_multiple_d_xdl_cshuffle.hpp
@@ -602,9 +602,7 @@ struct DeviceContractionMultipleD_Xdl_CShuffle
             return false;
         }
 
-        if(ck::get_device_name() != "gfx90a" && ck::get_device_name() != "gfx940" &&
-           ck::get_device_name() != "gfx941" && ck::get_device_name() != "gfx942" &&
-           std::is_same<ADataType, double>::value)
+        if(!ck::is_lds_direct_load_supported() && std::is_same<ADataType, double>::value)
         {
             return false;
         }
diff --git a/include/ck/tensor_operation/gpu/device/impl/device_elementwise_3d_impl.hpp b/include/ck/tensor_operation/gpu/device/impl/device_elementwise_3d_impl.hpp
index 67b6f87465..be778b7137 100644
--- a/include/ck/tensor_operation/gpu/device/impl/device_elementwise_3d_impl.hpp
+++ b/include/ck/tensor_operation/gpu/device/impl/device_elementwise_3d_impl.hpp
@@ -294,7 +294,7 @@ struct DeviceElementwise3dImpl : public DeviceElementwise<InDataTypeTuple,
     bool IsSupportedArgument(const BaseArgument* p_arg) override
     {
         if((ck::get_device_name() == "gfx940" || ck::get_device_name() == "gfx941" ||
-            ck::get_device_name() == "gfx942"))
+            ck::get_device_name() == "gfx942" || ck::get_device_name() == "gfx950"))
         {
             return false;
         }
diff --git a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_splitk_lds_direct_load.hpp b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_splitk_lds_direct_load.hpp
index 94306a4c95..67abb68d30 100644
--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_splitk_lds_direct_load.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_splitk_lds_direct_load.hpp
@@ -39,7 +39,7 @@ __global__ void
                                                   const CElementwiseOperation c_element_op)
 {
 #if(!defined(__HIP_DEVICE_COMPILE__) || defined(__gfx908__) || defined(__gfx90a__) || \
-    defined(__gfx940__) || defined(__gfx941__) || defined(__gfx942__))
+    defined(__gfx94__))
     constexpr index_t shared_size = GridwiseGemm::GetSharedMemoryNumberOfByte();
 
     __shared__ uint8_t p_shared[shared_size];
diff --git a/include/ck/utility/amd_xdlops.hpp b/include/ck/utility/amd_xdlops.hpp
index 0ee52b9570..4bc079dbb6 100644
--- a/include/ck/utility/amd_xdlops.hpp
+++ b/include/ck/utility/amd_xdlops.hpp
@@ -5,7 +5,7 @@
 
 namespace ck {
 // Define the common macro for MI300 models
-#if defined(__gfx940__) || defined(__gfx941__) || defined(__gfx942__)
+#if defined(__gfx940__) || defined(__gfx941__) || defined(__gfx942__) || defined(__gfx950__)
 #define __gfx94__
 #endif
 
diff --git a/include/ck/utility/data_type.hpp b/include/ck/utility/data_type.hpp
index 31ae71880a..5827d4d3f3 100644
--- a/include/ck/utility/data_type.hpp
+++ b/include/ck/utility/data_type.hpp
@@ -189,7 +189,7 @@ struct vector_type<T, 1>
     }
 };
 
-int static err = 0;
+__device__ int static err = 0;
 template <typename T>
 struct vector_type<T, 2>
 {
diff --git a/include/ck/utility/type_convert.hpp b/include/ck/utility/type_convert.hpp
index 6bbff98312..fa7aac04c7 100644
--- a/include/ck/utility/type_convert.hpp
+++ b/include/ck/utility/type_convert.hpp
@@ -9,7 +9,7 @@
 
 namespace ck {
 // Define the common macro for MI300 models
-#if defined(__gfx940__) || defined(__gfx941__) || defined(__gfx942__)
+#if defined(__gfx940__) || defined(__gfx941__) || defined(__gfx942__) || defined(__gfx950__)
 #define __gfx94__
 #endif
 
diff --git a/test/batched_gemm/CMakeLists.txt b/test/batched_gemm/CMakeLists.txt
index 9482821b68..1bb24f4c1a 100644
--- a/test/batched_gemm/CMakeLists.txt
+++ b/test/batched_gemm/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
@@ -6,4 +6,4 @@ foreach(gpu IN LISTS GPU_TARGETS)
    target_link_libraries(test_batched_gemm PRIVATE utility device_batched_gemm_instance)
    set(target 1)
  endif()
-endforeach()
\ No newline at end of file
+endforeach()
diff --git a/test/batched_gemm_gemm/CMakeLists.txt b/test/batched_gemm_gemm/CMakeLists.txt
index 03f1d3a4eb..ca14fcee04 100644
--- a/test/batched_gemm_gemm/CMakeLists.txt
+++ b/test/batched_gemm_gemm/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
@@ -10,4 +10,4 @@ foreach(gpu IN LISTS GPU_TARGETS)
       set(target 1)
     endif()
  endif()
-endforeach()
\ No newline at end of file
+endforeach()
diff --git a/test/batched_gemm_reduce/CMakeLists.txt b/test/batched_gemm_reduce/CMakeLists.txt
index 32c6ee85d1..24bfcf5ebc 100644
--- a/test/batched_gemm_reduce/CMakeLists.txt
+++ b/test/batched_gemm_reduce/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/test/batched_gemm_softmax_gemm/CMakeLists.txt b/test/batched_gemm_softmax_gemm/CMakeLists.txt
index c011a6a3c5..b6cd11f7c0 100644
--- a/test/batched_gemm_softmax_gemm/CMakeLists.txt
+++ b/test/batched_gemm_softmax_gemm/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
@@ -10,4 +10,4 @@ foreach(gpu IN LISTS GPU_TARGETS)
       set(target 1)
     endif()
  endif()
-endforeach()
\ No newline at end of file
+endforeach()
diff --git a/test/batched_gemm_softmax_gemm_permute/CMakeLists.txt b/test/batched_gemm_softmax_gemm_permute/CMakeLists.txt
index 3164863eef..05f2e855f3 100644
--- a/test/batched_gemm_softmax_gemm_permute/CMakeLists.txt
+++ b/test/batched_gemm_softmax_gemm_permute/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
@@ -26,4 +26,4 @@ foreach(gpu IN LISTS GPU_TARGETS)
     endif()
    set(target 1)
  endif()
-endforeach()
\ No newline at end of file
+endforeach()
diff --git a/test/contraction/CMakeLists.txt b/test/contraction/CMakeLists.txt
index a86e72fddb..6e647f02a0 100644
--- a/test/contraction/CMakeLists.txt
+++ b/test/contraction/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
     if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/test/convnd_bwd_data/CMakeLists.txt b/test/convnd_bwd_data/CMakeLists.txt
index f734b46f53..cca2a955c5 100644
--- a/test/convnd_bwd_data/CMakeLists.txt
+++ b/test/convnd_bwd_data/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
@@ -6,4 +6,4 @@ foreach(gpu IN LISTS GPU_TARGETS)
    target_link_libraries(test_convnd_bwd_data PRIVATE utility device_conv1d_bwd_data_instance device_conv2d_bwd_data_instance device_conv3d_bwd_data_instance)
    set(target 1)
  endif()
-endforeach()
\ No newline at end of file
+endforeach()
diff --git a/test/convnd_fwd/CMakeLists.txt b/test/convnd_fwd/CMakeLists.txt
index 745aceffc9..5a6c650d72 100644
--- a/test/convnd_fwd/CMakeLists.txt
+++ b/test/convnd_fwd/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/test/gemm_layernorm/CMakeLists.txt b/test/gemm_layernorm/CMakeLists.txt
index bfc4404bd8..a1403a5f71 100644
--- a/test/gemm_layernorm/CMakeLists.txt
+++ b/test/gemm_layernorm/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/test/gemm_split_k/CMakeLists.txt b/test/gemm_split_k/CMakeLists.txt
index caf30fca59..61cdf4ee94 100644
--- a/test/gemm_split_k/CMakeLists.txt
+++ b/test/gemm_split_k/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/test/grouped_convnd_bwd_weight/CMakeLists.txt b/test/grouped_convnd_bwd_weight/CMakeLists.txt
index b167943c97..5e6baa9933 100644
--- a/test/grouped_convnd_bwd_weight/CMakeLists.txt
+++ b/test/grouped_convnd_bwd_weight/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list_xdl gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list_xdl gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 list(APPEND gpu_list_wmma gfx1100 gfx1101 gfx1102)
 
 set(target 0)
@@ -17,4 +17,4 @@ foreach(gpu IN LISTS GPU_TARGETS)
    target_link_libraries(test_grouped_convnd_bwd_weight_interface PRIVATE utility)
    set(target 1)
  endif()
-endforeach()
\ No newline at end of file
+endforeach()
diff --git a/test/grouped_gemm/CMakeLists.txt b/test/grouped_gemm/CMakeLists.txt
index 8c57b667e2..a4b500f4af 100644
--- a/test/grouped_gemm/CMakeLists.txt
+++ b/test/grouped_gemm/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/test/transpose/CMakeLists.txt b/test/transpose/CMakeLists.txt
index 530cc9d72d..e288461c8b 100644
--- a/test/transpose/CMakeLists.txt
+++ b/test/transpose/CMakeLists.txt
@@ -1,4 +1,4 @@
-list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942)
+list(APPEND gpu_list gfx908 gfx90a gfx940 gfx941 gfx942 gfx950)
 set(target 0)
 foreach(gpu IN LISTS GPU_TARGETS)
  if(gpu IN_LIST gpu_list AND target EQUAL 0)
diff --git a/test/wrapper/CMakeLists.txt b/test/wrapper/CMakeLists.txt
index cadc146795..a12584c0a7 100644
--- a/test/wrapper/CMakeLists.txt
+++ b/test/wrapper/CMakeLists.txt
@@ -8,7 +8,7 @@ add_gtest_executable(test_partition test_partition.cpp)
 target_link_libraries(test_partition PRIVATE utility)
 if(GPU_TARGETS MATCHES "gfx908" OR GPU_TARGETS MATCHES "gfx90a" OR
    GPU_TARGETS MATCHES "gfx940" OR GPU_TARGETS MATCHES "gfx941" OR
-   GPU_TARGETS MATCHES "gfx942")
+   GPU_TARGETS MATCHES "gfx942" OR GPU_TARGETS MATCHES "gfx950")
     add_gtest_executable(test_gemm test_gemm.cpp)
     target_link_libraries(test_gemm PRIVATE utility)
 endif()