From 75fe8235f78520fbbfff7c9c035ecd9f1aa3e6f6 Mon Sep 17 00:00:00 2001 From: Cedric Nugteren Date: Sat, 23 Jul 2016 10:20:11 +0200 Subject: Improved the XgemvFastRot kernel by tiled loading of the input matrix A, enabling better memory performance --- src/routines/level2/xgemv.cpp | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'src/routines') diff --git a/src/routines/level2/xgemv.cpp b/src/routines/level2/xgemv.cpp index 2842ef07..e4d407c8 100644 --- a/src/routines/level2/xgemv.cpp +++ b/src/routines/level2/xgemv.cpp @@ -122,7 +122,7 @@ StatusCode Xgemv::MatVec(const Layout layout, const Transpose a_transpose, } if (fast_kernel_rot) { kernel_name = "XgemvFastRot"; - global_size = m_real / db_["WPT3"]; + global_size = m_real; local_size = db_["WGS3"]; } -- cgit v1.2.3