From 20da5e33a86eda746c17cbdb7bfd295d9f92f074 Mon Sep 17 00:00:00 2001 From: Cedric Nugteren Date: Wed, 6 Sep 2017 21:50:42 +0200 Subject: Split the database files over multiple directories and files; first step towards separate compilation --- src/database/kernels/transpose/transpose_64.hpp | 80 +++++++++++++++++++++++++ 1 file changed, 80 insertions(+) create mode 100644 src/database/kernels/transpose/transpose_64.hpp (limited to 'src/database/kernels/transpose/transpose_64.hpp') diff --git a/src/database/kernels/transpose/transpose_64.hpp b/src/database/kernels/transpose/transpose_64.hpp new file mode 100644 index 00000000..ddad3bac --- /dev/null +++ b/src/database/kernels/transpose/transpose_64.hpp @@ -0,0 +1,80 @@ + +// ================================================================================================= +// This file is part of the CLBlast project. The project is licensed under Apache Version 2.0. It +// is auto-generated by the 'scripts/database/database.py' Python script. +// +// This file populates the database with best-found tuning parameters for the 'Transpose64' kernels. +// +// ================================================================================================= + +namespace clblast { +namespace database { + +const DatabaseEntry TransposeDouble = { + "Transpose", Precision::kDouble, {"TRA_DIM", "TRA_PAD", "TRA_SHUFFLE", "TRA_WPT"}, { + { // AMD GPUs + kDeviceTypeGPU, "AMD", { + { "AMD Radeon R9 M370X Compute Engine", { 4, 0, 1, 4 } }, + { "Ellesmere", { 4, 0, 1, 4 } }, + { "Fiji", { 8, 1, 1, 2 } }, + { "Hawaii", { 16, 0, 1, 1 } }, + { "Oland", { 8, 1, 1, 2 } }, + { "Pitcairn", { 4, 0, 1, 2 } }, + { "Tahiti", { 4, 1, 1, 4 } }, + { "Tonga", { 4, 0, 1, 4 } }, + { "default", { 4, 0, 1, 4 } }, + } + }, + { // ARM GPUs + kDeviceTypeGPU, "ARM", { + { "Mali-T628", { 8, 0, 0, 1 } }, + { "default", { 8, 0, 0, 1 } }, + } + }, + { // Intel CPUs + kDeviceTypeCPU, "Intel", { + { "Intel(R) Core(TM) i7-2670QM CPU @ 2.20GHz", { 4, 1, 0, 8 } }, + { "Intel(R) Core(TM) i5-6200U CPU @ 2.30GHz", { 4, 0, 0, 4 } }, + { "Intel(R) Core(TM) i7 CPU 920 @ 2.67GHz", { 4, 1, 0, 4 } }, + { "Intel(R) Core(TM) i7-3770 CPU @ 3.40GHz", { 4, 1, 0, 8 } }, + { "Intel(R) Core(TM) i7-4790K CPU @ 4.00GHz", { 4, 0, 0, 16 } }, + { "Intel(R) Core(TM) i7-5930K CPU @ 3.50GHz", { 4, 0, 0, 8 } }, + { "default", { 4, 1, 0, 8 } }, + } + }, + { // Intel accelerators + kDeviceTypeAccelerator, "Intel", { + { "Intel(R) Many Integrated Core Acceleration Card", { 32, 1, 0, 1 } }, + { "default", { 32, 1, 0, 1 } }, + } + }, + { // NVIDIA GPUs + kDeviceTypeGPU, "NVIDIA", { + { "GRID K520", { 16, 1, 1, 2 } }, + { "GeForce GTX 1070", { 8, 0, 1, 2 } }, + { "GeForce GTX 1080", { 8, 0, 0, 2 } }, + { "GeForce GTX 480", { 8, 1, 0, 2 } }, + { "GeForce GTX 670", { 16, 1, 1, 2 } }, + { "GeForce GTX 680", { 16, 1, 1, 2 } }, + { "GeForce GTX 750", { 16, 1, 0, 1 } }, + { "GeForce GTX 750 Ti", { 32, 1, 0, 2 } }, + { "GeForce GTX 980", { 16, 0, 0, 2 } }, + { "GeForce GTX TITAN", { 8, 0, 0, 2 } }, + { "GeForce GTX TITAN Black", { 16, 1, 0, 2 } }, + { "GeForce GTX TITAN X", { 32, 1, 0, 1 } }, + { "TITAN X (Pascal)", { 16, 1, 0, 2 } }, + { "Tesla K20m", { 16, 1, 0, 2 } }, + { "Tesla K40m", { 16, 1, 1, 2 } }, + { "default", { 16, 1, 1, 2 } }, + } + }, + { // Default + kDeviceTypeAll, "default", { + { "default", { 16, 1, 1, 2 } }, + } + }, + } +}; + +} // namespace database +} // namespace clblast -- cgit v1.2.3