diff options
| -rw-r--r-- | CHANGELOG | 1 | ||||
| -rw-r--r-- | test/diagnostics.cpp | 3 |
2 files changed, 4 insertions, 0 deletions
@@ -4,6 +4,7 @@ Development (next version) - The tuning database now has a dictionary to translate vendor/device names to a common set - The tuners can now distinguish between different AMD GPU board names of the same architecture - The tuners can now use particle-swarm optimisation to search more efficiently (thanks to 'mcian') +- Improved performance for small problems on NVIDIA hardware by caching the device name - Further improved compilation time of database.cpp - Added a small diagnostics helper executable - Various minor fixes and enhancements diff --git a/test/diagnostics.cpp b/test/diagnostics.cpp index 6872ed6f..99b936f8 100644 --- a/test/diagnostics.cpp +++ b/test/diagnostics.cpp @@ -83,6 +83,9 @@ void OpenCLDiagnostics(int argc, char *argv[]) { printf("* queue.GetContext() %.4lf ms\n", TimeFunction(kNumRuns, [&](){queue.GetContext();} )); printf("* queue.GetDevice() %.4lf ms\n", TimeFunction(kNumRuns, [&](){queue.GetDevice();} )); printf("* device.Name() %.4lf ms\n", TimeFunction(kNumRuns, [&](){device.Name();} )); + printf("* device.Vendor() %.4lf ms\n", TimeFunction(kNumRuns, [&](){device.Vendor();} )); + printf("* device.Version() %.4lf ms\n", TimeFunction(kNumRuns, [&](){device.Version();} )); + printf("* device.Platform() %.4lf ms\n", TimeFunction(kNumRuns, [&](){device.Platform();} )); printf("* Buffer<float>(context, 1024) %.4lf ms\n", TimeFunction(kNumRuns, [&](){Buffer<float>(context, 1024);} )); printf("\n"); |