diff options
author | Gian Marco Iodice <gianmarco.iodice@arm.com> | 2018-04-19 12:05:08 +0100 |
---|---|---|
committer | Anthony Barbier <anthony.barbier@arm.com> | 2018-11-02 16:51:17 +0000 |
commit | bb36a8efc1092f66798e3b880c55ec488021bb02 (patch) | |
tree | 62e0265d84575bc10496c84f4908ed27529166ea /src/core/CL/CLKernelLibrary.cpp | |
parent | 4dcb583c052e14f08809cc9ee420e690264e7bbe (diff) | |
download | ComputeLibrary-bb36a8efc1092f66798e3b880c55ec488021bb02.tar.gz |
COMPMID-922 - CLGEMM FP16 optimizations - part2
This patch improves of ~30 % GEMM fp16 when the reshape is required
The results have been reported at the following confluence page:
https://confluence.arm.com/display/MLENG/GEMM+FP16+performance%3A+ACL+18.05
Change-Id: I8233095a7e9ab06f1f915782a25dd41653b49140
Reviewed-on: https://eu-gerrit-1.euhpc.arm.com/128254
Reviewed-by: Anthony Barbier <anthony.barbier@arm.com>
Tested-by: Jenkins <bsgcomp@arm.com>
Diffstat (limited to 'src/core/CL/CLKernelLibrary.cpp')
-rw-r--r-- | src/core/CL/CLKernelLibrary.cpp | 3 |
1 files changed, 2 insertions, 1 deletions
diff --git a/src/core/CL/CLKernelLibrary.cpp b/src/core/CL/CLKernelLibrary.cpp index 7e3eebc3b4..f1be935df3 100644 --- a/src/core/CL/CLKernelLibrary.cpp +++ b/src/core/CL/CLKernelLibrary.cpp @@ -230,7 +230,8 @@ const std::map<std::string, std::string> CLKernelLibrary::_kernel_program_map = { "gemm_mv", "gemv.cl" }, { "gemm_mv_quantized", "gemv.cl" }, { "gemm_mm_interleaved_transposed_f16", "gemm.cl" }, - { "gemm_mm_interleaved_transposed_f32_midgard", "gemm.cl" }, + { "gemm_mm_interleaved_transposed_f16_bifrost", "gemm.cl" }, + { "gemm_mm_interleaved_transposed_f32", "gemm.cl" }, { "gemm_mm_interleaved_transposed_f32_bifrost", "gemm.cl" }, { "gemm_mm_interleaved_transposed_qs8", "gemm.cl" }, { "gemm_mm_interleaved_transposed_qs16", "gemm.cl" }, |