diff --git a/driver/device_implicit_gemm_convolution_2_chwn_cyxk_khwn.hpp b/driver/device_implicit_gemm_convolution_2_chwn_cyxk_khwn.hpp
index bf7cdc8c5a..bab0b5d7fe 100644
--- a/driver/device_implicit_gemm_convolution_2_chwn_cyxk_khwn.hpp
+++ b/driver/device_implicit_gemm_convolution_2_chwn_cyxk_khwn.hpp
@@ -190,7 +190,7 @@ void device_implicit_gemm_convolution_2_chwn_cyxk_khwn(InDesc,
     constexpr index_t WeiBlockCopyDataPerRead = 4;
 
     constexpr index_t BlockSize = 256;
-#elif 1
+#elif 0
     // 1x1, 14x14, Vega 20, disable lds_double_buffer, enable register double buffer
     constexpr index_t BPerBlock = 64;
     constexpr index_t KPerBlock = 128;
diff --git a/src/include/gridwise_implicit_gemm_convolution_2_chwn_cyxk_khwn.hip.hpp b/src/include/gridwise_implicit_gemm_convolution_2_chwn_cyxk_khwn.hip.hpp
index fd223dfad4..657c233e5e 100644
--- a/src/include/gridwise_implicit_gemm_convolution_2_chwn_cyxk_khwn.hip.hpp
+++ b/src/include/gridwise_implicit_gemm_convolution_2_chwn_cyxk_khwn.hip.hpp
@@ -315,7 +315,7 @@ class gridwise_implicit_gemm_convolution_2_chwn_cyxk_khwn
                     auto f_accum = [](auto& acc, const auto&& v) { acc += v; };
 #if 0
                     blockwise_gemm.Run
-#elif 1
+#elif 0
                     blockwise_gemm.Run_RegisterDoubleBuffer
 #elif 1
                     blockwise_gemm.Run_asm