Skip to content

Test throughput on sample with lumi events

The following discussion from !950 (merged) should be addressed:

  • @dcampora started a discussion: (+4 comments)

    Depending on how time-consuming these kernels are, I would suggest to create instead a list of events that should pass this condition, to improve parallelism. But before doing that, please check the throughput and profile of these kernels to judge whether the added complexity is worth it.