From 08ee57f494648d8e64415dd8249797534f314308 Mon Sep 17 00:00:00 2001 From: Cedric Nugteren Date: Mon, 10 Oct 2016 16:41:41 +0200 Subject: Updated the tuning results for the GTX 750 Ti GPU --- src/database/kernels/copy.hpp | 10 +++++----- src/database/kernels/pad.hpp | 8 ++++---- src/database/kernels/padtranspose.hpp | 2 +- src/database/kernels/transpose.hpp | 2 +- src/database/kernels/xaxpy.hpp | 6 +++--- src/database/kernels/xdot.hpp | 4 ++++ src/database/kernels/xgemm.hpp | 8 ++++---- src/database/kernels/xgemm_direct.hpp | 12 ++++++------ src/database/kernels/xgemv.hpp | 6 +++--- src/database/kernels/xgemv_fast.hpp | 5 ++--- src/database/kernels/xgemv_fast_rot.hpp | 6 ++++-- src/database/kernels/xger.hpp | 8 ++++++-- 12 files changed, 43 insertions(+), 34 deletions(-) (limited to 'src') diff --git a/src/database/kernels/copy.hpp b/src/database/kernels/copy.hpp index a6b7dfe8..2eab308d 100644 --- a/src/database/kernels/copy.hpp +++ b/src/database/kernels/copy.hpp @@ -84,7 +84,7 @@ const Database::DatabaseEntry Database::CopySingle = { { "GeForce GTX 670", { {"COPY_DIMX",16}, {"COPY_DIMY",32}, {"COPY_VW",4}, {"COPY_WPT",1} } }, { "GeForce GTX 680", { {"COPY_DIMX",32}, {"COPY_DIMY",16}, {"COPY_VW",4}, {"COPY_WPT",1} } }, { "GeForce GTX 750", { {"COPY_DIMX",32}, {"COPY_DIMY",8}, {"COPY_VW",2}, {"COPY_WPT",2} } }, - { "GeForce GTX 750 Ti", { {"COPY_DIMX",32}, {"COPY_DIMY",8}, {"COPY_VW",2}, {"COPY_WPT",1} } }, + { "GeForce GTX 750 Ti", { {"COPY_DIMX",16}, {"COPY_DIMY",32}, {"COPY_VW",2}, {"COPY_WPT",2} } }, { "GeForce GTX 980", { {"COPY_DIMX",32}, {"COPY_DIMY",16}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "GeForce GTX TITAN", { {"COPY_DIMX",32}, {"COPY_DIMY",8}, {"COPY_VW",2}, {"COPY_WPT",4} } }, { "GeForce GTX TITAN X", { {"COPY_DIMX",32}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",2} } }, @@ -147,7 +147,7 @@ const Database::DatabaseEntry Database::CopyComplexSingle = { { "GeForce GTX 480", { {"COPY_DIMX",16}, {"COPY_DIMY",16}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "GeForce GTX 670", { {"COPY_DIMX",16}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "GeForce GTX 750", { {"COPY_DIMX",16}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",2} } }, - { "GeForce GTX 750 Ti", { {"COPY_DIMX",32}, {"COPY_DIMY",16}, {"COPY_VW",1}, {"COPY_WPT",1} } }, + { "GeForce GTX 750 Ti", { {"COPY_DIMX",16}, {"COPY_DIMY",32}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "GeForce GTX 980", { {"COPY_DIMX",8}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "GeForce GTX TITAN X", { {"COPY_DIMX",16}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "Tesla K20m", { {"COPY_DIMX",8}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",4} } }, @@ -205,13 +205,13 @@ const Database::DatabaseEntry Database::CopyDouble = { { "GeForce GTX 670", { {"COPY_DIMX",8}, {"COPY_DIMY",8}, {"COPY_VW",2}, {"COPY_WPT",1} } }, { "GeForce GTX 680", { {"COPY_DIMX",16}, {"COPY_DIMY",32}, {"COPY_VW",2}, {"COPY_WPT",1} } }, { "GeForce GTX 750", { {"COPY_DIMX",8}, {"COPY_DIMY",16}, {"COPY_VW",2}, {"COPY_WPT",1} } }, - { "GeForce GTX 750 Ti", { {"COPY_DIMX",8}, {"COPY_DIMY",8}, {"COPY_VW",2}, {"COPY_WPT",1} } }, + { "GeForce GTX 750 Ti", { {"COPY_DIMX",16}, {"COPY_DIMY",8}, {"COPY_VW",2}, {"COPY_WPT",1} } }, { "GeForce GTX 980", { {"COPY_DIMX",32}, {"COPY_DIMY",8}, {"COPY_VW",2}, {"COPY_WPT",1} } }, { "GeForce GTX TITAN", { {"COPY_DIMX",16}, {"COPY_DIMY",32}, {"COPY_VW",2}, {"COPY_WPT",2} } }, { "GeForce GTX TITAN X", { {"COPY_DIMX",32}, {"COPY_DIMY",16}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "Tesla K20m", { {"COPY_DIMX",8}, {"COPY_DIMY",8}, {"COPY_VW",2}, {"COPY_WPT",1} } }, { "Tesla K40m", { {"COPY_DIMX",8}, {"COPY_DIMY",8}, {"COPY_VW",2}, {"COPY_WPT",2} } }, - { "default", { {"COPY_DIMX",32}, {"COPY_DIMY",32}, {"COPY_VW",2}, {"COPY_WPT",1} } }, + { "default", { {"COPY_DIMX",32}, {"COPY_DIMY",16}, {"COPY_VW",2}, {"COPY_WPT",1} } }, } }, { // Default @@ -264,7 +264,7 @@ const Database::DatabaseEntry Database::CopyComplexDouble = { { "GeForce GTX 670", { {"COPY_DIMX",16}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "GeForce GTX 680", { {"COPY_DIMX",8}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "GeForce GTX 750", { {"COPY_DIMX",32}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",1} } }, - { "GeForce GTX 750 Ti", { {"COPY_DIMX",32}, {"COPY_DIMY",16}, {"COPY_VW",1}, {"COPY_WPT",1} } }, + { "GeForce GTX 750 Ti", { {"COPY_DIMX",16}, {"COPY_DIMY",16}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "GeForce GTX 980", { {"COPY_DIMX",8}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "GeForce GTX TITAN", { {"COPY_DIMX",16}, {"COPY_DIMY",16}, {"COPY_VW",1}, {"COPY_WPT",1} } }, { "GeForce GTX TITAN X", { {"COPY_DIMX",16}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",1} } }, diff --git a/src/database/kernels/pad.hpp b/src/database/kernels/pad.hpp index 3cfabaf4..49393a36 100644 --- a/src/database/kernels/pad.hpp +++ b/src/database/kernels/pad.hpp @@ -84,7 +84,7 @@ const Database::DatabaseEntry Database::PadSingle = { { "GeForce GTX 670", { {"PAD_DIMX",32}, {"PAD_DIMY",8}, {"PAD_WPTX",2}, {"PAD_WPTY",2} } }, { "GeForce GTX 680", { {"PAD_DIMX",16}, {"PAD_DIMY",8}, {"PAD_WPTX",4}, {"PAD_WPTY",1} } }, { "GeForce GTX 750", { {"PAD_DIMX",32}, {"PAD_DIMY",16}, {"PAD_WPTX",4}, {"PAD_WPTY",2} } }, - { "GeForce GTX 750 Ti", { {"PAD_DIMX",32}, {"PAD_DIMY",8}, {"PAD_WPTX",4}, {"PAD_WPTY",1} } }, + { "GeForce GTX 750 Ti", { {"PAD_DIMX",16}, {"PAD_DIMY",8}, {"PAD_WPTX",4}, {"PAD_WPTY",1} } }, { "GeForce GTX 980", { {"PAD_DIMX",16}, {"PAD_DIMY",8}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, { "GeForce GTX TITAN", { {"PAD_DIMX",32}, {"PAD_DIMY",8}, {"PAD_WPTX",2}, {"PAD_WPTY",1} } }, { "GeForce GTX TITAN X", { {"PAD_DIMX",16}, {"PAD_DIMY",16}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, @@ -154,13 +154,13 @@ const Database::DatabaseEntry Database::PadComplexSingle = { { "GeForce GTX 670", { {"PAD_DIMX",16}, {"PAD_DIMY",8}, {"PAD_WPTX",1}, {"PAD_WPTY",2} } }, { "GeForce GTX 680", { {"PAD_DIMX",16}, {"PAD_DIMY",32}, {"PAD_WPTX",1}, {"PAD_WPTY",2} } }, { "GeForce GTX 750", { {"PAD_DIMX",32}, {"PAD_DIMY",8}, {"PAD_WPTX",2}, {"PAD_WPTY",1} } }, - { "GeForce GTX 750 Ti", { {"PAD_DIMX",16}, {"PAD_DIMY",16}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, + { "GeForce GTX 750 Ti", { {"PAD_DIMX",16}, {"PAD_DIMY",8}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, { "GeForce GTX 980", { {"PAD_DIMX",16}, {"PAD_DIMY",16}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, { "GeForce GTX TITAN", { {"PAD_DIMX",16}, {"PAD_DIMY",8}, {"PAD_WPTX",2}, {"PAD_WPTY",1} } }, { "GeForce GTX TITAN X", { {"PAD_DIMX",16}, {"PAD_DIMY",8}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, { "Tesla K20m", { {"PAD_DIMX",32}, {"PAD_DIMY",8}, {"PAD_WPTX",1}, {"PAD_WPTY",2} } }, { "Tesla K40m", { {"PAD_DIMX",16}, {"PAD_DIMY",8}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, - { "default", { {"PAD_DIMX",16}, {"PAD_DIMY",8}, {"PAD_WPTX",2}, {"PAD_WPTY",1} } }, + { "default", { {"PAD_DIMX",16}, {"PAD_DIMY",8}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, } }, { // Default @@ -272,7 +272,7 @@ const Database::DatabaseEntry Database::PadComplexDouble = { { "GeForce GTX 670", { {"PAD_DIMX",32}, {"PAD_DIMY",8}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, { "GeForce GTX 680", { {"PAD_DIMX",8}, {"PAD_DIMY",8}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, { "GeForce GTX 750", { {"PAD_DIMX",8}, {"PAD_DIMY",8}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, - { "GeForce GTX 750 Ti", { {"PAD_DIMX",32}, {"PAD_DIMY",32}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, + { "GeForce GTX 750 Ti", { {"PAD_DIMX",16}, {"PAD_DIMY",32}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, { "GeForce GTX 980", { {"PAD_DIMX",16}, {"PAD_DIMY",16}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, { "GeForce GTX TITAN", { {"PAD_DIMX",8}, {"PAD_DIMY",32}, {"PAD_WPTX",1}, {"PAD_WPTY",2} } }, { "GeForce GTX TITAN X", { {"PAD_DIMX",16}, {"PAD_DIMY",8}, {"PAD_WPTX",1}, {"PAD_WPTY",1} } }, diff --git a/src/database/kernels/padtranspose.hpp b/src/database/kernels/padtranspose.hpp index 88bd4ea7..61951b6f 100644 --- a/src/database/kernels/padtranspose.hpp +++ b/src/database/kernels/padtranspose.hpp @@ -272,7 +272,7 @@ const Database::DatabaseEntry Database::PadtransposeComplexDouble = { { "GeForce GTX 670", { {"PADTRA_PAD",1}, {"PADTRA_TILE",16}, {"PADTRA_WPT",1} } }, { "GeForce GTX 680", { {"PADTRA_PAD",1}, {"PADTRA_TILE",32}, {"PADTRA_WPT",1} } }, { "GeForce GTX 750", { {"PADTRA_PAD",1}, {"PADTRA_TILE",16}, {"PADTRA_WPT",1} } }, - { "GeForce GTX 750 Ti", { {"PADTRA_PAD",1}, {"PADTRA_TILE",16}, {"PADTRA_WPT",2} } }, + { "GeForce GTX 750 Ti", { {"PADTRA_PAD",1}, {"PADTRA_TILE",8}, {"PADTRA_WPT",2} } }, { "GeForce GTX 980", { {"PADTRA_PAD",0}, {"PADTRA_TILE",16}, {"PADTRA_WPT",1} } }, { "GeForce GTX TITAN", { {"PADTRA_PAD",1}, {"PADTRA_TILE",16}, {"PADTRA_WPT",1} } }, { "GeForce GTX TITAN X", { {"PADTRA_PAD",1}, {"PADTRA_TILE",32}, {"PADTRA_WPT",1} } }, diff --git a/src/database/kernels/transpose.hpp b/src/database/kernels/transpose.hpp index 0e1b608e..7619878e 100644 --- a/src/database/kernels/transpose.hpp +++ b/src/database/kernels/transpose.hpp @@ -207,7 +207,7 @@ const Database::DatabaseEntry Database::TransposeDouble = { { "GeForce GTX 670", { {"TRA_DIM",16}, {"TRA_PAD",1}, {"TRA_SHUFFLE",1}, {"TRA_WPT",2} } }, { "GeForce GTX 680", { {"TRA_DIM",16}, {"TRA_PAD",1}, {"TRA_SHUFFLE",1}, {"TRA_WPT",2} } }, { "GeForce GTX 750", { {"TRA_DIM",16}, {"TRA_PAD",1}, {"TRA_SHUFFLE",0}, {"TRA_WPT",1} } }, - { "GeForce GTX 750 Ti", { {"TRA_DIM",32}, {"TRA_PAD",1}, {"TRA_SHUFFLE",0}, {"TRA_WPT",1} } }, + { "GeForce GTX 750 Ti", { {"TRA_DIM",32}, {"TRA_PAD",1}, {"TRA_SHUFFLE",0}, {"TRA_WPT",2} } }, { "GeForce GTX 980", { {"TRA_DIM",16}, {"TRA_PAD",0}, {"TRA_SHUFFLE",0}, {"TRA_WPT",2} } }, { "GeForce GTX TITAN", { {"TRA_DIM",8}, {"TRA_PAD",0}, {"TRA_SHUFFLE",0}, {"TRA_WPT",2} } }, { "GeForce GTX TITAN X", { {"TRA_DIM",32}, {"TRA_PAD",1}, {"TRA_SHUFFLE",0}, {"TRA_WPT",1} } }, diff --git a/src/database/kernels/xaxpy.hpp b/src/database/kernels/xaxpy.hpp index 9c1bcd99..89c7f48a 100644 --- a/src/database/kernels/xaxpy.hpp +++ b/src/database/kernels/xaxpy.hpp @@ -84,7 +84,7 @@ const Database::DatabaseEntry Database::XaxpySingle = { { "GeForce GTX 670", { {"VW",2}, {"WGS",64}, {"WPT",1} } }, { "GeForce GTX 680", { {"VW",1}, {"WGS",128}, {"WPT",1} } }, { "GeForce GTX 750", { {"VW",1}, {"WGS",64}, {"WPT",1} } }, - { "GeForce GTX 750 Ti", { {"VW",1}, {"WGS",1024}, {"WPT",1} } }, + { "GeForce GTX 750 Ti", { {"VW",2}, {"WGS",64}, {"WPT",1} } }, { "GeForce GTX 980", { {"VW",1}, {"WGS",1024}, {"WPT",1} } }, { "GeForce GTX TITAN", { {"VW",4}, {"WGS",256}, {"WPT",1} } }, { "GeForce GTX TITAN X", { {"VW",1}, {"WGS",64}, {"WPT",1} } }, @@ -213,7 +213,7 @@ const Database::DatabaseEntry Database::XaxpyDouble = { { "GeForce GTX 670", { {"VW",1}, {"WGS",64}, {"WPT",1} } }, { "GeForce GTX 680", { {"VW",1}, {"WGS",64}, {"WPT",1} } }, { "GeForce GTX 750", { {"VW",1}, {"WGS",128}, {"WPT",1} } }, - { "GeForce GTX 750 Ti", { {"VW",1}, {"WGS",64}, {"WPT",1} } }, + { "GeForce GTX 750 Ti", { {"VW",1}, {"WGS",256}, {"WPT",2} } }, { "GeForce GTX 980", { {"VW",1}, {"WGS",256}, {"WPT",1} } }, { "GeForce GTX TITAN", { {"VW",2}, {"WGS",1024}, {"WPT",1} } }, { "GeForce GTX TITAN X", { {"VW",1}, {"WGS",512}, {"WPT",1} } }, @@ -272,7 +272,7 @@ const Database::DatabaseEntry Database::XaxpyComplexDouble = { { "GeForce GTX 670", { {"VW",1}, {"WGS",256}, {"WPT",1} } }, { "GeForce GTX 680", { {"VW",1}, {"WGS",64}, {"WPT",1} } }, { "GeForce GTX 750", { {"VW",1}, {"WGS",1024}, {"WPT",1} } }, - { "GeForce GTX 750 Ti", { {"VW",1}, {"WGS",256}, {"WPT",2} } }, + { "GeForce GTX 750 Ti", { {"VW",1}, {"WGS",64}, {"WPT",2} } }, { "GeForce GTX 980", { {"VW",1}, {"WGS",1024}, {"WPT",1} } }, { "GeForce GTX TITAN", { {"VW",1}, {"WGS",64}, {"WPT",4} } }, { "GeForce GTX TITAN X", { {"VW",1}, {"WGS",1024}, {"WPT",1} } }, diff --git a/src/database/kernels/xdot.hpp b/src/database/kernels/xdot.hpp index 987a990d..51ca2cd7 100644 --- a/src/database/kernels/xdot.hpp +++ b/src/database/kernels/xdot.hpp @@ -68,6 +68,7 @@ const Database::DatabaseEntry Database::XdotSingle = { { "GeForce GTX 670", { {"WGS1",512}, {"WGS2",1024} } }, { "GeForce GTX 680", { {"WGS1",128}, {"WGS2",128} } }, { "GeForce GTX 750", { {"WGS1",128}, {"WGS2",32} } }, + { "GeForce GTX 750 Ti", { {"WGS1",64}, {"WGS2",32} } }, { "GeForce GTX 980", { {"WGS1",256}, {"WGS2",32} } }, { "GeForce GTX TITAN X", { {"WGS1",256}, {"WGS2",32} } }, { "Tesla K20m", { {"WGS1",1024}, {"WGS2",32} } }, @@ -119,6 +120,7 @@ const Database::DatabaseEntry Database::XdotComplexSingle = { { "GeForce GTX 670", { {"WGS1",256}, {"WGS2",32} } }, { "GeForce GTX 680", { {"WGS1",128}, {"WGS2",64} } }, { "GeForce GTX 750", { {"WGS1",64}, {"WGS2",32} } }, + { "GeForce GTX 750 Ti", { {"WGS1",64}, {"WGS2",32} } }, { "GeForce GTX 980", { {"WGS1",256}, {"WGS2",64} } }, { "GeForce GTX TITAN X", { {"WGS1",256}, {"WGS2",32} } }, { "Tesla K20m", { {"WGS1",512}, {"WGS2",32} } }, @@ -160,6 +162,7 @@ const Database::DatabaseEntry Database::XdotDouble = { { "GeForce GTX 670", { {"WGS1",256}, {"WGS2",32} } }, { "GeForce GTX 680", { {"WGS1",128}, {"WGS2",64} } }, { "GeForce GTX 750", { {"WGS1",64}, {"WGS2",256} } }, + { "GeForce GTX 750 Ti", { {"WGS1",128}, {"WGS2",64} } }, { "GeForce GTX 980", { {"WGS1",128}, {"WGS2",32} } }, { "GeForce GTX TITAN X", { {"WGS1",256}, {"WGS2",32} } }, { "Tesla K20m", { {"WGS1",512}, {"WGS2",32} } }, @@ -201,6 +204,7 @@ const Database::DatabaseEntry Database::XdotComplexDouble = { { "GeForce GTX 670", { {"WGS1",512}, {"WGS2",128} } }, { "GeForce GTX 680", { {"WGS1",256}, {"WGS2",64} } }, { "GeForce GTX 750", { {"WGS1",256}, {"WGS2",32} } }, + { "GeForce GTX 750 Ti", { {"WGS1",64}, {"WGS2",32} } }, { "GeForce GTX 980", { {"WGS1",64}, {"WGS2",32} } }, { "GeForce GTX TITAN X", { {"WGS1",128}, {"WGS2",32} } }, { "Tesla K20m", { {"WGS1",128}, {"WGS2",32} } }, diff --git a/src/database/kernels/xgemm.hpp b/src/database/kernels/xgemm.hpp index e289c542..5532ef6b 100644 --- a/src/database/kernels/xgemm.hpp +++ b/src/database/kernels/xgemm.hpp @@ -77,7 +77,7 @@ const Database::DatabaseEntry Database::XgemmSingle = { { "GeForce GTX 670", { {"KWG",16}, {"KWI",2}, {"MDIMA",8}, {"MDIMC",8}, {"MWG",64}, {"NDIMB",16}, {"NDIMC",16}, {"NWG",64}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",0}, {"VWM",2}, {"VWN",4} } }, { "GeForce GTX 680", { {"KWG",32}, {"KWI",8}, {"MDIMA",8}, {"MDIMC",16}, {"MWG",64}, {"NDIMB",32}, {"NDIMC",16}, {"NWG",128}, {"SA",1}, {"SB",1}, {"STRM",0}, {"STRN",0}, {"VWM",4}, {"VWN",2} } }, { "GeForce GTX 750", { {"KWG",16}, {"KWI",2}, {"MDIMA",16}, {"MDIMC",16}, {"MWG",64}, {"NDIMB",32}, {"NDIMC",8}, {"NWG",128}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",1}, {"VWM",1}, {"VWN",2} } }, - { "GeForce GTX 750 Ti", { {"KWG",16}, {"KWI",2}, {"MDIMA",32}, {"MDIMC",16}, {"MWG",128}, {"NDIMB",16}, {"NDIMC",8}, {"NWG",128}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",1}, {"VWM",4}, {"VWN",4} } }, + { "GeForce GTX 750 Ti", { {"KWG",16}, {"KWI",2}, {"MDIMA",16}, {"MDIMC",16}, {"MWG",128}, {"NDIMB",32}, {"NDIMC",8}, {"NWG",64}, {"SA",1}, {"SB",1}, {"STRM",0}, {"STRN",1}, {"VWM",8}, {"VWN",2} } }, { "GeForce GTX 980", { {"KWG",16}, {"KWI",2}, {"MDIMA",16}, {"MDIMC",16}, {"MWG",64}, {"NDIMB",16}, {"NDIMC",8}, {"NWG",128}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",0}, {"VWM",4}, {"VWN",8} } }, { "GeForce GTX TITAN", { {"KWG",16}, {"KWI",8}, {"MDIMA",32}, {"MDIMC",16}, {"MWG",64}, {"NDIMB",8}, {"NDIMC",8}, {"NWG",64}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",0}, {"VWM",2}, {"VWN",2} } }, { "GeForce GTX TITAN X", { {"KWG",16}, {"KWI",2}, {"MDIMA",8}, {"MDIMC",16}, {"MWG",128}, {"NDIMB",8}, {"NDIMC",8}, {"NWG",128}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",1}, {"VWM",4}, {"VWN",8} } }, @@ -147,7 +147,7 @@ const Database::DatabaseEntry Database::XgemmComplexSingle = { { "GeForce GTX 670", { {"KWG",16}, {"KWI",2}, {"MDIMA",32}, {"MDIMC",32}, {"MWG",64}, {"NDIMB",32}, {"NDIMC",8}, {"NWG",32}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",1}, {"VWM",1}, {"VWN",1} } }, { "GeForce GTX 680", { {"KWG",16}, {"KWI",2}, {"MDIMA",32}, {"MDIMC",16}, {"MWG",64}, {"NDIMB",32}, {"NDIMC",32}, {"NWG",128}, {"SA",1}, {"SB",0}, {"STRM",0}, {"STRN",0}, {"VWM",2}, {"VWN",2} } }, { "GeForce GTX 750", { {"KWG",16}, {"KWI",8}, {"MDIMA",16}, {"MDIMC",16}, {"MWG",64}, {"NDIMB",16}, {"NDIMC",16}, {"NWG",64}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",0}, {"VWM",2}, {"VWN",2} } }, - { "GeForce GTX 750 Ti", { {"KWG",16}, {"KWI",2}, {"MDIMA",32}, {"MDIMC",8}, {"MWG",128}, {"NDIMB",16}, {"NDIMC",32}, {"NWG",128}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",0}, {"VWM",1}, {"VWN",4} } }, + { "GeForce GTX 750 Ti", { {"KWG",16}, {"KWI",2}, {"MDIMA",16}, {"MDIMC",8}, {"MWG",32}, {"NDIMB",32}, {"NDIMC",16}, {"NWG",64}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",0}, {"VWM",1}, {"VWN",2} } }, { "GeForce GTX 980", { {"KWG",32}, {"KWI",8}, {"MDIMA",32}, {"MDIMC",32}, {"MWG",64}, {"NDIMB",16}, {"NDIMC",16}, {"NWG",64}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",0}, {"VWM",2}, {"VWN",1} } }, { "GeForce GTX TITAN", { {"KWG",16}, {"KWI",8}, {"MDIMA",16}, {"MDIMC",16}, {"MWG",64}, {"NDIMB",32}, {"NDIMC",16}, {"NWG",64}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",0}, {"VWM",1}, {"VWN",1} } }, { "GeForce GTX TITAN X", { {"KWG",16}, {"KWI",2}, {"MDIMA",8}, {"MDIMC",8}, {"MWG",64}, {"NDIMB",8}, {"NDIMC",8}, {"NWG",32}, {"SA",1}, {"SB",0}, {"STRM",1}, {"STRN",1}, {"VWM",1}, {"VWN",4} } }, @@ -206,7 +206,7 @@ const Database::DatabaseEntry Database::XgemmDouble = { { "GeForce GTX 670", { {"KWG",32}, {"KWI",8}, {"MDIMA",16}, {"MDIMC",32}, {"MWG",128}, {"NDIMB",16}, {"NDIMC",8}, {"NWG",32}, {"SA",0}, {"SB",1}, {"STRM",1}, {"STRN",0}, {"VWM",1}, {"VWN",1} } }, { "GeForce GTX 680", { {"KWG",32}, {"KWI",8}, {"MDIMA",8}, {"MDIMC",8}, {"MWG",32}, {"NDIMB",16}, {"NDIMC",32}, {"NWG",128}, {"SA",1}, {"SB",0}, {"STRM",0}, {"STRN",1}, {"VWM",2}, {"VWN",4} } }, { "GeForce GTX 750", { {"KWG",32}, {"KWI",8}, {"MDIMA",16}, {"MDIMC",32}, {"MWG",64}, {"NDIMB",16}, {"NDIMC",8}, {"NWG",128}, {"SA",0}, {"SB",0}, {"STRM",0}, {"STRN",1}, {"VWM",2}, {"VWN",1} } }, - { "GeForce GTX 750 Ti", { {"KWG",16}, {"KWI",2}, {"MDIMA",16}, {"MDIMC",8}, {"MWG",32}, {"NDIMB",16}, {"NDIMC",32}, {"NWG",32}, {"SA",0}, {"SB",0}, {"STRM",1}, {"STRN",1}, {"VWM",2}, {"VWN",1} } }, + { "GeForce GTX 750 Ti", { {"KWG",32}, {"KWI",2}, {"MDIMA",8}, {"MDIMC",8}, {"MWG",32}, {"NDIMB",16}, {"NDIMC",16}, {"NWG",32}, {"SA",0}, {"SB",0}, {"STRM",0}, {"STRN",0}, {"VWM",4}, {"VWN",2} } }, { "GeForce GTX 980", { {"KWG",32}, {"KWI",8}, {"MDIMA",16}, {"MDIMC",8}, {"MWG",64}, {"NDIMB",32}, {"NDIMC",32}, {"NWG",128}, {"SA",0}, {"SB",0}, {"STRM",1}, {"STRN",0}, {"VWM",2}, {"VWN",4} } }, { "GeForce GTX TITAN", { {"KWG",16}, {"KWI",8}, {"MDIMA",16}, {"MDIMC",8}, {"MWG",32}, {"NDIMB",16}, {"NDIMC",32}, {"NWG",128}, {"SA",1}, {"SB",1}, {"STRM",1}, {"STRN",1}, {"VWM",2}, {"VWN",2} } }, { "GeForce GTX TITAN X", { {"KWG",16}, {"KWI",8}, {"MDIMA",16}, {"MDIMC",16}, {"MWG",16}, {"NDIMB",16}, {"NDIMC",16}, {"NWG",16}, {"SA",0}, {"SB",0}, {"STRM",0}, {"STRN",0}, {"VWM",1}, {"VWN",1} } }, @@ -265,7 +265,7 @@ const Database::DatabaseEntry Database::XgemmComplexDouble = { { "GeForce GTX 670", { {"KWG",32}, {"KWI",8}, {"MDIMA",16}, {"MDIMC",8}, {"MWG",16}, {"NDIMB",16}, {"NDIMC",32}, {"NWG",64}, {"SA",1}, {"SB",0}, {"STRM",0}, {"STRN",1}, {"VWM",1}, {"VWN",2} } }, { "GeForce GTX 680", { {"KWG",16}, {"KWI",8}, {"MDIMA",16}, {"MDIMC",8}, {"MWG",64}, {"NDIMB",16}, {"NDIMC",32}, {"NWG",32}, {"SA",0}, {"SB",1}, {"STRM",1}, {"STRN",0}, {"VWM",1}, {"VWN",1} } }, { "GeForce GTX 750", { {"KWG",32}, {"KWI",2}, {"MDIMA",8}, {"MDIMC",32}, {"MWG",32}, {"NDIMB",8}, {"NDIMC",8}, {"NWG",64}, {"SA",0}, {"SB",0}, {"STRM",1}, {"STRN",0}, {"VWM",1}, {"VWN",4} } }, - { "GeForce GTX 750 Ti", { {"KWG",16}, {"KWI",2}, {"MDIMA",16}, {"MDIMC",16}, {"MWG",16}, {"NDIMB",8}, {"NDIMC",8}, {"NWG",32}, {"SA",0}, {"SB",0}, {"STRM",0}, {"STRN",0}, {"VWM",1}, {"VWN",4} } }, + { "GeForce GTX 750 Ti", { {"KWG",32}, {"KWI",2}, {"MDIMA",8}, {"MDIMC",8}, {"MWG",16}, {"NDIMB",8}, {"NDIMC",8}, {"NWG",32}, {"SA",0}, {"SB",0}, {"STRM",0}, {"STRN",0}, {"VWM",1}, {"VWN",1} } }, { "GeForce GTX 980", { {"KWG",16}, {"KWI",2}, {"MDIMA",16}, {"MDIMC",8}, {"MWG",32}, {"NDIMB",8}, {"NDIMC",16}, {"NWG",128}, {"SA",0}, {"SB",0}, {"STRM",1}, {"STRN",1}, {"VWM",2}, {"VWN",2} } }, { "GeForce GTX TITAN X", { {"KWG",32}, {"KWI",8}, {"MDIMA",16}, {"MDIMC",16}, {"MWG",128}, {"NDIMB",16}, {"NDIMC",16}, {"NWG",32}, {"SA",0}, {"SB",0}, {"STRM",1}, {"STRN",0}, {"VWM",1}, {"VWN",1} } }, { "Tesla K20m", { {"KWG",32}, {"KWI",2}, {"MDIMA",32}, {"MDIMC",8}, {"MWG",32}, {"NDIMB",16}, {"NDIMC",16}, {"NWG",64}, {"SA",0}, {"SB",0}, {"STRM",1}, {"STRN",0}, {"VWM",1}, {"VWN",1} } }, diff --git a/src/database/kernels/xgemm_direct.hpp b/src/database/kernels/xgemm_direct.hpp index 202deb1f..53dfeaa6 100644 --- a/src/database/kernels/xgemm_direct.hpp +++ b/src/database/kernels/xgemm_direct.hpp @@ -42,8 +42,8 @@ const Database::DatabaseEntry Database::XgemmDirectSingle = { }, { // NVIDIA GPUs kDeviceTypeGPU, "NVIDIA", { - { "GeForce GTX 750 Ti", { {"KWID",2}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",8}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",4}, {"VWND",4}, {"WGD",32} } }, - { "default", { {"KWID",2}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",8}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",4}, {"VWND",4}, {"WGD",32} } }, + { "GeForce GTX 750 Ti", { {"KWID",2}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",8}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",4}, {"VWND",2}, {"WGD",32} } }, + { "default", { {"KWID",2}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",8}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",4}, {"VWND",2}, {"WGD",32} } }, } }, { // Default @@ -72,8 +72,8 @@ const Database::DatabaseEntry Database::XgemmDirectComplexSingle = { }, { // NVIDIA GPUs kDeviceTypeGPU, "NVIDIA", { - { "GeForce GTX 750 Ti", { {"KWID",2}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",8}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",1}, {"VWND",2}, {"WGD",16} } }, - { "default", { {"KWID",2}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",8}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",1}, {"VWND",2}, {"WGD",16} } }, + { "GeForce GTX 750 Ti", { {"KWID",16}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",16}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",2}, {"VWND",1}, {"WGD",16} } }, + { "default", { {"KWID",16}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",16}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",2}, {"VWND",1}, {"WGD",16} } }, } }, { // Default @@ -96,8 +96,8 @@ const Database::DatabaseEntry Database::XgemmDirectDouble = { }, { // NVIDIA GPUs kDeviceTypeGPU, "NVIDIA", { - { "GeForce GTX 750 Ti", { {"KWID",2}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",8}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",2}, {"VWND",2}, {"WGD",32} } }, - { "default", { {"KWID",2}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",8}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",2}, {"VWND",2}, {"WGD",32} } }, + { "GeForce GTX 750 Ti", { {"KWID",2}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",8}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",2}, {"VWND",4}, {"WGD",32} } }, + { "default", { {"KWID",2}, {"MDIMAD",8}, {"MDIMCD",8}, {"NDIMBD",8}, {"NDIMCD",8}, {"PADA",1}, {"PADB",1}, {"VWMD",2}, {"VWND",4}, {"WGD",32} } }, } }, { // Default diff --git a/src/database/kernels/xgemv.hpp b/src/database/kernels/xgemv.hpp index e5e8845e..a4a339dc 100644 --- a/src/database/kernels/xgemv.hpp +++ b/src/database/kernels/xgemv.hpp @@ -77,7 +77,7 @@ const Database::DatabaseEntry Database::XgemvSingle = { { "GeForce GTX 670", { {"WGS1",64}, {"WPT1",1} } }, { "GeForce GTX 680", { {"WGS1",256}, {"WPT1",1} } }, { "GeForce GTX 750", { {"WGS1",256}, {"WPT1",1} } }, - { "GeForce GTX 750 Ti", { {"WGS1",256}, {"WPT1",1} } }, + { "GeForce GTX 750 Ti", { {"WGS1",32}, {"WPT1",1} } }, { "GeForce GTX 980", { {"WGS1",128}, {"WPT1",1} } }, { "GeForce GTX TITAN", { {"WGS1",256}, {"WPT1",1} } }, { "GeForce GTX TITAN X", { {"WGS1",256}, {"WPT1",1} } }, @@ -140,7 +140,7 @@ const Database::DatabaseEntry Database::XgemvComplexSingle = { { "GeForce GTX 670", { {"WGS1",64}, {"WPT1",1} } }, { "GeForce GTX 680", { {"WGS1",64}, {"WPT1",1} } }, { "GeForce GTX 750", { {"WGS1",128}, {"WPT1",1} } }, - { "GeForce GTX 750 Ti", { {"WGS1",64}, {"WPT1",1} } }, + { "GeForce GTX 750 Ti", { {"WGS1",32}, {"WPT1",1} } }, { "GeForce GTX TITAN", { {"WGS1",256}, {"WPT1",1} } }, { "default", { {"WGS1",64}, {"WPT1",1} } }, } @@ -188,7 +188,7 @@ const Database::DatabaseEntry Database::XgemvDouble = { { "GeForce GTX 670", { {"WGS1",128}, {"WPT1",1} } }, { "GeForce GTX 680", { {"WGS1",128}, {"WPT1",1} } }, { "GeForce GTX 750", { {"WGS1",64}, {"WPT1",1} } }, - { "GeForce GTX 750 Ti", { {"WGS1",64}, {"WPT1",1} } }, + { "GeForce GTX 750 Ti", { {"WGS1",32}, {"WPT1",1} } }, { "GeForce GTX 980", { {"WGS1",64}, {"WPT1",1} } }, { "GeForce GTX TITAN", { {"WGS1",256}, {"WPT1",1} } }, { "GeForce GTX TITAN X", { {"WGS1",64}, {"WPT1",1} } }, diff --git a/src/database/kernels/xgemv_fast.hpp b/src/database/kernels/xgemv_fast.hpp index 52af628c..7e5afaae 100644 --- a/src/database/kernels/xgemv_fast.hpp +++ b/src/database/kernels/xgemv_fast.hpp @@ -77,7 +77,7 @@ const Database::DatabaseEntry Database::XgemvFastSingle = { { "GeForce GTX 670", { {"VW2",2}, {"WGS2",256}, {"WPT2",2} } }, { "GeForce GTX 680", { {"VW2",1}, {"WGS2",128}, {"WPT2",1} } }, { "GeForce GTX 750", { {"VW2",1}, {"WGS2",256}, {"WPT2",1} } }, - { "GeForce GTX 750 Ti", { {"VW2",1}, {"WGS2",256}, {"WPT2",1} } }, + { "GeForce GTX 750 Ti", { {"VW2",2}, {"WGS2",32}, {"WPT2",2} } }, { "GeForce GTX 980", { {"VW2",1}, {"WGS2",256}, {"WPT2",1} } }, { "GeForce GTX TITAN", { {"VW2",1}, {"WGS2",256}, {"WPT2",1} } }, { "GeForce GTX TITAN X", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } }, @@ -139,7 +139,6 @@ const Database::DatabaseEntry Database::XgemvFastComplexSingle = { { "GeForce GTX 480", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } }, { "GeForce GTX 670", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } }, { "GeForce GTX 680", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } }, - { "GeForce GTX 750 Ti", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } }, { "default", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } }, } }, @@ -186,7 +185,7 @@ const Database::DatabaseEntry Database::XgemvFastDouble = { { "GeForce GTX 670", { {"VW2",1}, {"WGS2",128}, {"WPT2",1} } }, { "GeForce GTX 680", { {"VW2",1}, {"WGS2",128}, {"WPT2",1} } }, { "GeForce GTX 750", { {"VW2",2}, {"WGS2",256}, {"WPT2",2} } }, - { "GeForce GTX 750 Ti", { {"VW2",1}, {"WGS2",256}, {"WPT2",1} } }, + { "GeForce GTX 750 Ti", { {"VW2",1}, {"WGS2",32}, {"WPT2",2} } }, { "GeForce GTX 980", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } }, { "GeForce GTX TITAN", { {"VW2",1}, {"WGS2",256}, {"WPT2",1} } }, { "GeForce GTX TITAN X", { {"VW2",1}, {"WGS2",128}, {"WPT2",1} } }, diff --git a/src/database/kernels/xgemv_fast_rot.hpp b/src/database/kernels/xgemv_fast_rot.hpp index 328094e1..58854189 100644 --- a/src/database/kernels/xgemv_fast_rot.hpp +++ b/src/database/kernels/xgemv_fast_rot.hpp @@ -51,8 +51,9 @@ const Database::DatabaseEntry Database::XgemvFastRotSingle = { }, { // NVIDIA GPUs kDeviceTypeGPU, "NVIDIA", { + { "GeForce GTX 750 Ti", { {"VW3",8}, {"WGS3",32}, {"WPT3",32} } }, { "GeForce GTX TITAN", { {"VW3",1}, {"WGS3",16}, {"WPT3",16} } }, - { "default", { {"VW3",1}, {"WGS3",16}, {"WPT3",16} } }, + { "default", { {"VW3",8}, {"WGS3",32}, {"WPT3",32} } }, } }, { // Default @@ -114,8 +115,9 @@ const Database::DatabaseEntry Database::XgemvFastRotDouble = { }, { // NVIDIA GPUs kDeviceTypeGPU, "NVIDIA", { + { "GeForce GTX 750 Ti", { {"VW3",4}, {"WGS3",32}, {"WPT3",16} } }, { "GeForce GTX TITAN", { {"VW3",1}, {"WGS3",16}, {"WPT3",16} } }, - { "default", { {"VW3",1}, {"WGS3",16}, {"WPT3",16} } }, + { "default", { {"VW3",4}, {"WGS3",32}, {"WPT3",16} } }, } }, { // Default diff --git a/src/database/kernels/xger.hpp b/src/database/kernels/xger.hpp index 3e9c25c1..d057c65c 100644 --- a/src/database/kernels/xger.hpp +++ b/src/database/kernels/xger.hpp @@ -76,8 +76,9 @@ const Database::DatabaseEntry Database::XgerSingle = { { "GeForce GTX 670", { {"WGS1",32}, {"WGS2",8}, {"WPT",2} } }, { "GeForce GTX 680", { {"WGS1",128}, {"WGS2",1}, {"WPT",4} } }, { "GeForce GTX 750", { {"WGS1",64}, {"WGS2",16}, {"WPT",4} } }, + { "GeForce GTX 750 Ti", { {"WGS1",64}, {"WGS2",1}, {"WPT",2} } }, { "GeForce GTX TITAN", { {"WGS1",32}, {"WGS2",4}, {"WPT",2} } }, - { "default", { {"WGS1",256}, {"WGS2",1}, {"WPT",4} } }, + { "default", { {"WGS1",128}, {"WGS2",1}, {"WPT",2} } }, } }, { // Default @@ -133,6 +134,7 @@ const Database::DatabaseEntry Database::XgerComplexSingle = { { "GeForce GTX 670", { {"WGS1",16}, {"WGS2",32}, {"WPT",2} } }, { "GeForce GTX 680", { {"WGS1",32}, {"WGS2",4}, {"WPT",2} } }, { "GeForce GTX 750", { {"WGS1",32}, {"WGS2",16}, {"WPT",4} } }, + { "GeForce GTX 750 Ti", { {"WGS1",32}, {"WGS2",8}, {"WPT",2} } }, { "GeForce GTX TITAN", { {"WGS1",16}, {"WGS2",16}, {"WPT",2} } }, { "default", { {"WGS1",64}, {"WGS2",2}, {"WPT",2} } }, } @@ -180,8 +182,9 @@ const Database::DatabaseEntry Database::XgerDouble = { { "GeForce GTX 670", { {"WGS1",32}, {"WGS2",32}, {"WPT",2} } }, { "GeForce GTX 680", { {"WGS1",128}, {"WGS2",4}, {"WPT",2} } }, { "GeForce GTX 750", { {"WGS1",256}, {"WGS2",2}, {"WPT",2} } }, + { "GeForce GTX 750 Ti", { {"WGS1",32}, {"WGS2",16}, {"WPT",1} } }, { "GeForce GTX TITAN", { {"WGS1",16}, {"WGS2",8}, {"WPT",2} } }, - { "default", { {"WGS1",256}, {"WGS2",2}, {"WPT",2} } }, + { "default", { {"WGS1",64}, {"WGS2",2}, {"WPT",2} } }, } }, { // Default @@ -227,6 +230,7 @@ const Database::DatabaseEntry Database::XgerComplexDouble = { { "GeForce GTX 670", { {"WGS1",8}, {"WGS2",16}, {"WPT",2} } }, { "GeForce GTX 680", { {"WGS1",8}, {"WGS2",16}, {"WPT",1} } }, { "GeForce GTX 750", { {"WGS1",8}, {"WGS2",32}, {"WPT",4} } }, + { "GeForce GTX 750 Ti", { {"WGS1",32}, {"WGS2",8}, {"WPT",2} } }, { "GeForce GTX TITAN", { {"WGS1",32}, {"WGS2",4}, {"WPT",2} } }, { "default", { {"WGS1",16}, {"WGS2",8}, {"WPT",2} } }, } -- cgit v1.2.3