f3959203 · Use _mm512_movepi32_mask to extract sign mask in AVX512 microkernels · Dec 23, 2020

1504db60 · Internal config change · Jun 08, 2020

9fefe013 · Cache blocking on last-level cache · May 23, 2020

739746d1 · Enable direct convolution path · May 15, 2020

35e3bba7 · foo · Mar 21, 2020

74427b5f · Update Raspberry Pi performance · Feb 13, 2020

a84e40bd · Pass proper CXX flags to built-in randomized models · Dec 11, 2019

d7337390 · !!!DO NOT MERGE!!! Disable FMA3 code path on x86 · Dec 06, 2019

191e5cd2 · Fix typo in CMakeLists · Nov 28, 2019

40a672f8 · Move generated micro-kernels into a subdirectory · Nov 25, 2019

e3fad191 · Fix wrong vreinterpret intrinsics in NEONFMA SpCHW DWCONV micro-kernels · Nov 22, 2019

10a02153 · Support bazel build w/ --config=chromiumos_x86_64. · Nov 22, 2019

df06d80b · Neon shuffle GEMM and IGEMM kernels. · Nov 20, 2019

ffd68401 · SSE2 exp and expminus evaluation stubs · Nov 15, 2019

346a9e58 · Sigmoid evaluation stubs, micro-kernels, and operator · Nov 15, 2019

797a8fea · NEONFMA versions of vector expf function · Nov 14, 2019

c76065f8 · Fix CMake build of elementwise microkernels · Nov 14, 2019

0ecc2ab6 · 4x8 GEMM for Cortex A53 · Nov 14, 2019

5abe43cc · ST1 post increment for Cortex A53 GEMM/IGEMM microkernels · Nov 13, 2019

88148155 · Internal config change · Oct 30, 2019