From ae0fc8612dba6faebf58c3ebbfae8d6e639d432d Mon Sep 17 00:00:00 2001 From: Georgios Pinitas Date: Mon, 30 Sep 2019 12:39:40 +0100 Subject: COMPMID-2452: Dot product optimizations on merge/transforms -Adds optimized gemm transforms for AArch64. -Optimized gemm merger to only support alpha==1 and (beta==0 || beta=1) cases Change-Id: I55793b12a0381f4fd53f521d0e57416809904d96 Signed-off-by: Georgios Pinitas Reviewed-on: https://review.mlplatform.org/c/2003 Tested-by: Arm Jenkins Reviewed-by: Michalis Spyrou --- src/core/NEON/kernels/arm_gemm/transforms/list.hpp | 1 + 1 file changed, 1 insertion(+) (limited to 'src/core/NEON/kernels/arm_gemm/transforms/list.hpp') diff --git a/src/core/NEON/kernels/arm_gemm/transforms/list.hpp b/src/core/NEON/kernels/arm_gemm/transforms/list.hpp index 9cd5983ce0..c0c2ca19d7 100644 --- a/src/core/NEON/kernels/arm_gemm/transforms/list.hpp +++ b/src/core/NEON/kernels/arm_gemm/transforms/list.hpp @@ -26,6 +26,7 @@ #include "a64_block16_interleave4_8bit.hpp" #include "a64_interleave_8way_16bit.hpp" #include "a64_interleave_8way_32bit.hpp" +#include "a64_interleave_8way_block4_8bit.hpp" #include "a64_interleave_8way_half_to_float.hpp" #include "a64_transpose_interleave_12way_16bit.hpp" #include "a64_transpose_interleave_12way_half_to_float.hpp" -- cgit v1.2.1