To give you some background “hyperthreading”, that is the ability for a single physical core to run more threads than 1, is turned off on the cluster by default because it does not work well with most HPC codes.
So the “maximum” threads a node in the standard partition can have is 36 since there Continue Reading »