Jon Holba
Jon Holba
tested some on a RHEL8 node. time seems to always give correct numbers. the weird thing is that mpirun with -np 1 or 2 restricts openmp from running on any...
adding `--bind-to core` when using -np >= 3 will give the same behavior as -np 1 or 2 adding `--bind-to none` will give the same behavior as -np >= 3...
This seems to be a bit slower than current main. My bigpoly test (edited to fail halfway through the forward model) goes from 1m1s in main to 1m16s in this...
Are we sending full snapshots every time now? The major reason for the slowdown is the parsing of dates. It went from ~51k date parses in main to ~436k date...
Getting #8310 merged would bring the difference between main and this PR down to 3-5 seconds in my test. Should still figure out why there is so much extra parsing...
#8304 may also help with performance