Skip to content
Snippets Groups Projects
Select Git revision
  • master default protected
1 result
You can move around the graph by using the arrow keys.
Created with Raphaël 2.2.022Jul211Jun4317May151430Apr2625241816121185432129Mar211528Feb26252220191815141312111087654131Jan302928242322211816111098721Dec18171413121110754328Nov26232221151413931Oct30262524232219181615121110854228Sep2721201917141312629Aug282724232221201615223Jul20191064319Jun151488May426Apr252423191817Wild draft for a sum factorized point diagonal preconditionerCode generation time evaluationof products and sumsCode generation time evaluation of matrix inversesAdd a show_options entry pointCorrect permuting_horizontal_add for Vec8dAdd permuting_lower/upper stuffMore fixupFixupHorizontal add mit opcoutner[!322] Register PDELab finite elements into UFL[!305] Performance transformation: Loop reordering in sumfact kernelPass number of threads as environment variableAdd threading to autotune targetRename autotune benchmark template filesFix pep8Reproduce old benchmarks through template file[skip ci] Simulate cache misses in autotune benchmarks[skip ci] Create autotune benchmark from template fileRemove option attributes from option arraysDefault to Null in schemeRead default options from scheme and fix some ini file errorsuse function for ssc macroAdd comment to option fileCorrectly load yaml files using pkg_resourcesValidate options using cerberusFix pep8 and add tests for fastdg3d and dg2dAdd tests for loop reordering[skip ci] Rewrite loop reordering[skip ci] Only precompute argument of horizontal add if necessary[skip ci] Non fastdg tests for loop reordering[skip ci] Logging for performance transformationsenable sde for instrumentation level >= 3adds function to get ssc marks[!324] Cmake documentation improvements[skip ci] Finish direct accumulation loop reordering[skip ci] Remove Opcounter from signature for autotuningcache basis evaluation in constructor[skip ci] Turn off opcounting for autotune generation[skip ci][WIP] Rewrite loop transformation with direct accumulation[skip ci] Always precompute the horizontal add

Datenschutz und Impressum