Skip to content

Removed many parallel constructions which slowed down the performance on AARTFAAC-12 data.

Duncan Kampert requested to merge padre-parallel into master
  • Sped up parallel loops by removing them, as the workload was too low to be parallelized.
  • Added -cpu flag handling in main.
  • Added OpenMP to project to attempt to parallelize heavy-workload loops.
  • Added baseline computation from antennas to remove need for lookup-table in some cases for slight speedups.
  • Removed qualitystatistics computation, as this is only necessary when running aoquality afterwards.
    • This will later be replaced by in-af2ms computation of qualitystatistics by means of gpu parallelization.
Edited by Duncan Kampert

Merge request reports