From 6f7b1e48e62fb14b79f2a2b293b018c4b197bdb7 Mon Sep 17 00:00:00 2001 From: Yermalayeu Ihar Date: Thu, 26 Dec 2024 17:25:53 +0300 Subject: [PATCH] =?UTF-8?q?*improve=20Performance=20of=20SynetConvolution3?= =?UTF-8?q?2f=20(NHWC,=20srcC=3D1,=20dst=D0=A1=3D1).?= MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit --- docs/2025.html | 1 + src/Simd/SimdAvx512bwSynetConvolution32fDirectNchw.cpp | 2 ++ src/Simd/SimdSse41SynetConvolution32fDirectNchw.cpp | 2 ++ src/Test/TestSynetConvolution32f.cpp | 7 +++++-- 4 files changed, 10 insertions(+), 2 deletions(-) diff --git a/docs/2025.html b/docs/2025.html index f9989315ef..fb5c7d73d0 100644 --- a/docs/2025.html +++ b/docs/2025.html @@ -55,6 +55,7 @@
Improving
  • Extend using of AMX-BF16 optimization of function DepthwiseConvolution_k7p3d1s1w8.
  • Extend using of AVX-512BW optimization of function Convolution32fNhwcDepthwise_k7p3d1s1w4.
  • Extend using of AMX-BF16 optimization of function DepthwiseConvolution_k5p2d1s1w8.
  • +
  • Performance of SynetConvolution32f (NHWC, srcC=1, dstŠ”=1).
  • Bug fixing