summaryrefslogtreecommitdiff
path: root/src/database/kernels
diff options
context:
space:
mode:
authorCedric Nugteren <web@cedricnugteren.nl>2016-09-10 14:01:21 +0200
committerCedric Nugteren <web@cedricnugteren.nl>2016-09-10 14:01:21 +0200
commit94163970ae2e4fec2e86de9640a546fbeee23324 (patch)
treefd92fb20b3ddee3b24438068d36902e893219248 /src/database/kernels
parenta2f83507033a20b534099c7b21d4a7466108e949 (diff)
parente21f32bc9928f87a8d0ff15797e2ed2ab65ded58 (diff)
Merge branch 'xgemm_tuner_exhaustive' into development
Diffstat (limited to 'src/database/kernels')
-rw-r--r--src/database/kernels/copy.hpp2
-rw-r--r--src/database/kernels/xaxpy.hpp8
-rw-r--r--src/database/kernels/xgemm.hpp2
-rw-r--r--src/database/kernels/xgemv.hpp2
-rw-r--r--src/database/kernels/xgemv_fast.hpp8
5 files changed, 11 insertions, 11 deletions
diff --git a/src/database/kernels/copy.hpp b/src/database/kernels/copy.hpp
index dc2011fd..a6b7dfe8 100644
--- a/src/database/kernels/copy.hpp
+++ b/src/database/kernels/copy.hpp
@@ -87,7 +87,7 @@ const Database::DatabaseEntry Database::CopySingle = {
{ "GeForce GTX 750 Ti", { {"COPY_DIMX",32}, {"COPY_DIMY",8}, {"COPY_VW",2}, {"COPY_WPT",1} } },
{ "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",2}, {"COPY_WPT",1} } },
+ { "GeForce GTX TITAN X", { {"COPY_DIMX",32}, {"COPY_DIMY",8}, {"COPY_VW",1}, {"COPY_WPT",2} } },
{ "Tesla K20m", { {"COPY_DIMX",8}, {"COPY_DIMY",8}, {"COPY_VW",4}, {"COPY_WPT",4} } },
{ "Tesla K40m", { {"COPY_DIMX",8}, {"COPY_DIMY",8}, {"COPY_VW",4}, {"COPY_WPT",2} } },
{ "default", { {"COPY_DIMX",8}, {"COPY_DIMY",16}, {"COPY_VW",4}, {"COPY_WPT",1} } },
diff --git a/src/database/kernels/xaxpy.hpp b/src/database/kernels/xaxpy.hpp
index 60fa7555..6e84ca5a 100644
--- a/src/database/kernels/xaxpy.hpp
+++ b/src/database/kernels/xaxpy.hpp
@@ -79,10 +79,10 @@ const Database::DatabaseEntry Database::XaxpySingle = {
{ // NVIDIA GPUs
kDeviceTypeGPU, "NVIDIA", {
{ "GRID K520", { {"VW",2}, {"WGS",64}, {"WPT",1} } },
- { "GeForce GTX 1070", { {"VW",2}, {"WGS",64}, {"WPT",1} } },
- { "GeForce GTX 480", { {"VW",4}, {"WGS",64}, {"WPT",1} } },
+ { "GeForce GTX 1070", { {"VW",1}, {"WGS",64}, {"WPT",4} } },
+ { "GeForce GTX 480", { {"VW",2}, {"WGS",128}, {"WPT",1} } },
{ "GeForce GTX 670", { {"VW",2}, {"WGS",64}, {"WPT",1} } },
- { "GeForce GTX 680", { {"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 980", { {"VW",1}, {"WGS",1024}, {"WPT",1} } },
@@ -209,7 +209,7 @@ const Database::DatabaseEntry Database::XaxpyDouble = {
kDeviceTypeGPU, "NVIDIA", {
{ "GRID K520", { {"VW",1}, {"WGS",64}, {"WPT",1} } },
{ "GeForce GTX 1070", { {"VW",1}, {"WGS",64}, {"WPT",8} } },
- { "GeForce GTX 480", { {"VW",2}, {"WGS",64}, {"WPT",1} } },
+ { "GeForce GTX 480", { {"VW",1}, {"WGS",128}, {"WPT",1} } },
{ "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} } },
diff --git a/src/database/kernels/xgemm.hpp b/src/database/kernels/xgemm.hpp
index c960592d..cc81cf6a 100644
--- a/src/database/kernels/xgemm.hpp
+++ b/src/database/kernels/xgemm.hpp
@@ -158,7 +158,7 @@ const Database::DatabaseEntry Database::XgemmDouble = {
"Xgemm", Precision::kDouble, {
{ // AMD GPUs
kDeviceTypeGPU, "AMD", {
- { "AMD Radeon R9 M370X Compute Engine", { {"KWG",32}, {"KWI",2}, {"MDIMA",32}, {"MDIMC",32}, {"MWG",64}, {"NDIMB",8}, {"NDIMC",8}, {"NWG",64}, {"SA",0}, {"SB",0}, {"STRM",1}, {"STRN",1}, {"VWM",2}, {"VWN",8} } },
+ { "AMD Radeon R9 M370X Compute Engine", { {"KWG",32}, {"KWI",2}, {"MDIMA",16}, {"MDIMC",16}, {"MWG",64}, {"NDIMB",8}, {"NDIMC",8}, {"NWG",32}, {"SA",0}, {"SB",0}, {"STRM",0}, {"STRN",0}, {"VWM",4}, {"VWN",4} } },
{ "Hawaii", { {"KWG",16}, {"KWI",8}, {"MDIMA",32}, {"MDIMC",8}, {"MWG",128}, {"NDIMB",8}, {"NDIMC",8}, {"NWG",32}, {"SA",0}, {"SB",1}, {"STRM",0}, {"STRN",0}, {"VWM",1}, {"VWN",4} } },
{ "Oland", { {"KWG",16}, {"KWI",2}, {"MDIMA",8}, {"MDIMC",16}, {"MWG",64}, {"NDIMB",16}, {"NDIMC",8}, {"NWG",16}, {"SA",0}, {"SB",0}, {"STRM",1}, {"STRN",1}, {"VWM",1}, {"VWN",1} } },
{ "Pitcairn", { {"KWG",32}, {"KWI",2}, {"MDIMA",32}, {"MDIMC",16}, {"MWG",64}, {"NDIMB",8}, {"NDIMC",16}, {"NWG",32}, {"SA",0}, {"SB",0}, {"STRM",0}, {"STRN",0}, {"VWM",1}, {"VWN",2} } },
diff --git a/src/database/kernels/xgemv.hpp b/src/database/kernels/xgemv.hpp
index 7e8e64e3..03e84525 100644
--- a/src/database/kernels/xgemv.hpp
+++ b/src/database/kernels/xgemv.hpp
@@ -88,7 +88,7 @@ const Database::DatabaseEntry Database::XgemvSingle = {
},
{ // Default
kDeviceTypeAll, "default", {
- { "default", { {"WGS1",64}, {"WPT1",4} } },
+ { "default", { {"WGS1",64}, {"WPT1",1} } },
}
},
}
diff --git a/src/database/kernels/xgemv_fast.hpp b/src/database/kernels/xgemv_fast.hpp
index f5e3e630..c12fcdca 100644
--- a/src/database/kernels/xgemv_fast.hpp
+++ b/src/database/kernels/xgemv_fast.hpp
@@ -60,7 +60,7 @@ const Database::DatabaseEntry Database::XgemvFastSingle = {
{ "Intel(R) HD Graphics Skylake ULT GT2", { {"VW2",1}, {"WGS2",256}, {"WPT2",1} } },
{ "Iris", { {"VW2",1}, {"WGS2",128}, {"WPT2",2} } },
{ "Iris Pro", { {"VW2",1}, {"WGS2",128}, {"WPT2",2} } },
- { "default", { {"VW2",1}, {"WGS2",32}, {"WPT2",1} } },
+ { "default", { {"VW2",2}, {"WGS2",64}, {"WPT2",2} } },
}
},
{ // Intel accelerators
@@ -88,7 +88,7 @@ const Database::DatabaseEntry Database::XgemvFastSingle = {
},
{ // Default
kDeviceTypeAll, "default", {
- { "default", { {"VW2",4}, {"WGS2",128}, {"WPT2",4} } },
+ { "default", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } },
}
},
}
@@ -123,7 +123,7 @@ const Database::DatabaseEntry Database::XgemvFastComplexSingle = {
{ "Intel(R) HD Graphics Skylake ULT GT2", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } },
{ "Iris", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } },
{ "Iris Pro", { {"VW2",4}, {"WGS2",128}, {"WPT2",4} } },
- { "default", { {"VW2",1}, {"WGS2",32}, {"WPT2",1} } },
+ { "default", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } },
}
},
{ // Intel accelerators
@@ -145,7 +145,7 @@ const Database::DatabaseEntry Database::XgemvFastComplexSingle = {
},
{ // Default
kDeviceTypeAll, "default", {
- { "default", { {"VW2",1}, {"WGS2",32}, {"WPT2",1} } },
+ { "default", { {"VW2",1}, {"WGS2",64}, {"WPT2",1} } },
}
},
}