projects / platform / core / ml / nnfw.git / commit
? search:
summary | shortlog | log | commit | commitdiff | tree
(parent: 0a5fb5b) | patch
Making comparison_op_quantized.cl generic with VEC_SIZE (#4042)
authorPrasanna R/SNAP /SRI-Bangalore/Engineer/삼성전자 <prasanna.r@samsung.com>
Tue, 18 Dec 2018 04:47:17 +0000 (10:17 +0530)
committer오형석/On-Device Lab(SR)/Staff Engineer/삼성전자 <hseok82.oh@samsung.com>
Tue, 18 Dec 2018 04:47:17 +0000 (13:47 +0900)
commit5a1045348d9b9d90886bac72a404e3185486439c
tree5d484acc32999f4136dfb4661d238cde01d843a7tree | snapshot
parent0a5fb5bbf07d5d306d94aa226a7172a97f932441commit | diff
Making comparison_op_quantized.cl generic with VEC_SIZE (#4042)

This patch makes comparison op quantized.cl more generic based on VEC_SIZE
replacement from 16 -> vec_size
This is done for better scaling of this kernel.

Signed-off-by: prasannar <prasanna.r@samsung.com>
libs/ARMComputeEx/src/core/CL/CLKernelLibrary.cpp diff | blob | history
libs/ARMComputeEx/src/core/CL/cl_kernels/comparison_op_quantized.cl diff | blob | history
libs/ARMComputeEx/src/core/CL/kernels/CLComparisonOpKernel.cpp diff | blob | history
Domain: Machine Learning / ML Framework;