- Nov 21, 2021
-
-
Jakub Klinkovský authored
Taken from https://github.com/jedbrown/cmake-modules
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
Updated some old benchmark scripts See merge request !109
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
The tnl-matrix-convert and tnl-grid-view tools are not part of TNL anymore. Both scripts were moved to the tnl-legacy repository.
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
Otherwise random things may break, since new nvcc versions use C++17 by default...
-
- Nov 20, 2021
-
-
Jakub Klinkovský authored
- the old v1.10.x does not build with Clang 13 - prevent installing GTest along with TNL, it is used only for tests that are not installed themselves
-
- Nov 17, 2021
-
-
Jakub Klinkovský authored
Benchmarks refactoring Closes #96 and #62 See merge request !107
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
Benchmarks: removed method newBenchmark and refactored logging of remaining metadata into table columns The global metadata like hardware information is logged into a separate JSON file, so the handling of MetadataMap in the Logging classes is not needed anymore.
-
Jakub Klinkovský authored
Fixes #96
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
Benchmarks: refactored logging to write directly into a file instead of an intermediate stringstream
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
- Nov 15, 2021
-
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
Fixes #62
-
Jakub Klinkovský authored
Moved common headers from Benchmarks to TNL/Benchmarks so they can be used in other projects more easily
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
The performance on 16 cores is about the same as before, but it takes significantly less tasks so it should be much faster on less cores.
-
Jakub Klinkovský authored
The build time improved from 4:48 to 1:43 on 16 cores, so the efficiency is not great, but it's still better than nothing. Maybe the benchmarks could be reorganized into more independent tasks that could be explicitly instantiated more efficiently (even if there would be less explicit instantiations eventually).
-
Jakub Klinkovský authored
Instead of writing the whole log as a valid JSON object, it is much easier to write one JSON object per line. This can be processed in Python with the Pandas package using the following code: df = pandas.read_json(open("foo.log", "r"), orient="records", lines=True) Also removed the outputMode parameter from the Benchmark and Logging classes, since it was pretty useless.
-
Jakub Klinkovský authored
-
Jakub Klinkovský authored
-