Closed ranjanan closed 3 years ago
Would be nice to run a small benchmark to make sure the performance didn't drop for a single thread.
Say I have 4 cores, and hence 4 threads, but I spawn 100 solves. Won't those 100 solves be concurrently updating this data structure, unless you schedule the solves to ensure that each processor is only working on one solve at a time?
Sorry I accidentally clicked merge. I have reverted it, but you may need to open a new PR. Hope that's ok. I do think that this PR does not fix the issue.
Let's continue this discussion in the upstream circuitscape repo.
Create
nthreads()
scratch spaces.