diff --git a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_streamk.hpp b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_streamk.hpp
index 4408b34870..70abcac0b1 100644
--- a/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_streamk.hpp
+++ b/include/ck/tensor_operation/gpu/grid/gridwise_gemm_xdlops_streamk.hpp
@@ -37,7 +37,8 @@ __global__ void
                                    index_t StrideC,
                                    typename GridwiseGemm::Block2CTileMap block_mapping)
 {
-#if(!defined(__HIP_DEVICE_COMPILE__) || defined(__gfx908__) || defined(__gfx90a__))
+#if(!defined(__HIP_DEVICE_COMPILE__) || defined(__gfx908__) || defined(__gfx90a__) || \
+    defined(__gfx940__) || defined(__gfx941__) || defined(__gfx942__))
     constexpr index_t shared_size = GridwiseGemm::GetSharedMemoryNumberOfByte();
 
     __shared__ uint8_t p_shared[shared_size];