Skip to content
Snippets Groups Projects
Select Git revision
  • 57-mutex-global
  • 66-install-skript-ins-repo-und-aufraumen
  • 82-yl_tmp-re-initialization
  • ParallelComputation_w_reduce
  • cmpxchgmutextest
  • first-optimization-attempts-fix-attempt
  • master default protected
  • mutex-global-pacxx-log-level-5
  • presentation-and-report
  • scatterkernelreorder-optimization
  • sequential_gridoperator
  • sequential_gridoperator_rebased
12 results
You can move around the graph by using the arrow keys.
Created with Raphaël 2.2.019Jun1817161211765432129May2827252421201918171615141312117653230Apr2926241810543228Mar2726252221201917161514131211109875no error on agamemnonnum_vertices_ now contains the actual number of verticesMerge branch 'gridoperator-device-v2' into 'master'Corrected previous compilation errorscorrected some of the previous errorsmoved scattering from host to devicenow uploads inverse connectivity informationvectors for z and w now know their size beforehandAdded device-v2 solutionReset to the master versionPrepared files for v2 of the gridoperatorMerge branch 'master' into first-optimization-attemptsCorrected errors... benchmark sais: this optimized version is slowerMerge branch 'auto-timer-names' into 'master'Don't demangle the whole function, just the trafo argumentFix bench to extract the correct timingsReplace id and scale lambdas by proper functorsGet rid of the CRTP in the gridoperatorChanged memory ordermoved around some loading operationsadded print statements to the test, but this doesnt shed much light on the issue. [ci skip]cmpxchgmutextestcmpxchgmutextestMerge branch 'show-mode-for-bench-skript' into 'master'Mention print-only mode in READMEGive the bench skript a print-only modeMerge branch 'default-benchmarking-setup' into 'master'Adjust Readme.bench: do all relevant executors by defaultMerge branch 'fix-bench-agamemnon-case-sensitivity' into 'master'Merge branch 'sufficiently-large-benchmark' into 'master'Benchmark: fix case sensitivity when checking hostnameDevelop a sufficiently large benchmark problemMerge branch 'cleanup-unused-gpuquadrature' into 'master'LOP: get rid of GPUQuadratureMerge branch 'benchmarking-setup' into 'master'Short readme on how to benchmarkMerge branch 'Unittest_rework' into 'master'Update GPUQuadratureRuleTest.hhMerge branch 'master' into 'Unittest_rework'Timings: output relative deviation"Samples" -> "samples" when outputting timings for consistency