Implement QuantizeLinear with avx512 #6260
yufenglee
force pushed
from
5dd0e819
5 years ago
yufenglee
force pushed
to
5d56f9f6
5 years ago
yufenglee
force pushed
to
1aed8a58
5 years ago
yufenglee
force pushed
to
c02f54f6
5 years ago
tracysh
dismissed these changes
on 2021-01-30
QuantizeLinear AVX512
1911c20e
add unit test and benchmark
a94737b3
Tune avx512 for small size
0f33cc04
refine code style
2b04720c
test intrin functions when determing compiler suppport of avx512
29908492
declare MlasQuantizeLinearKernel in quantize_avx512f
d4b46b2d
explicitly specialize MlasQuantizeLinearKernel
cd482b4b
split intrinsics and asm source files
98101d05
exclude avx512 instrinsics if compiler doesn't support
60c3a3c8
implement with Avx512 intrinsics for N < 16
0e4015e6
only run QuantizeLienar benchmark for avx512 machine
ba2a2a4b
old compiler doesn't support _cvtu32_mask16
0f2bbece
move avx512core check to checker helper
514592de
remove redundant code
9ba967a7
use _mm512_mask_cvtepi32_storeu_epi8
ecddff7f
remove the benchmark of quantizelinear
693561db
use _mm512_maskz_loadu_ps
5586e16c
yufenglee
dismissed their stale review
via 5586e16c
5 years ago
yufenglee
force pushed
from
0610dba1
to
5586e16c
5 years ago
tracysh
approved these changes
on 2021-02-01
yufenglee
merged
7264a067
into master 5 years ago
yufenglee
deleted the yufeng/quantize_linear branch 5 years ago
Assignees
No one assigned
Login to write a write a comment.
Login via GitHub