Skip to content
Snippets Groups Projects

Repository graph

You can move around the graph by using the arrow keys.
Select Git revision
  • cpu
  • hzdr-gpu-test
  • master default protected
  • multi-gpu
  • web
  • 2021-04-12
  • 2021-02-08
7 results
Created with Raphaël 2.2.012Apr22Mar191817168Feb12Aug327May16Mar25Apr13Nov17Feb8Oct2Sep124Aug21May10Apr966Mar15Feb10921Nov925Oct2416Bump version to 2021-04-122021-04-12 master2021-04-12 masterfix host-to-device transfer of POIsCI: Add tests for GPU runnerhzdr-gpu-testhzdr-gpu-testMerge branch 'ci-and-tests' into 'master'CI: Add pycodestyle and shellcheck to test stageTests: Also check POI data outputsTests: Add documentation about how to run the testsTests: Use GDAL for comparing the resultsAdd GitLab CI configurationTests: Use expected results for comparing and split CPU and GPU testsTests: Add files for expected resultsTests: Fix pycodestyle linting issuesUpdate version2021-02-082021-02-08Update .gitignorerefactor header file usageminor CUDA code tuningsAdd compatibility flags to nvcc commandFix no data value for arrival times outputFix bug in OpenMP parallelization.multi-gpumulti-gpuFix bug in OpenMP parallelization.cpucpuFix bug in OpenMP parallelization.webwebRevert "Print calculated bounding box"Add .gitignorePrint calculated bounding boxFix bug in OpenMP parallelization.replaced old NVFLAGS, should build for all platformsUpdate README.mdAdd READMEAdd licenseMoved svn branches to git branches, moved trunk to masterMoved svn branch cpu to git branchMoved svn branch multi-gpu to git branchMoved svn branch web to git branchFixed double free caused by multiple GPUs.Keep kernel data small by separating POI related data in a dedicated class. This was necessary to make the code run on the older Tesla C1060.Time series for POIs are now stored on GPU side and only copied back at the end of the simulation. This saves a lot of runtime especially for a large number of POIs.Write output file to disk before printing last progress.Added dist-packages for revision 34.Fixed time measurement in multi-threaded version.Added support for OpenMP within the build system.