# ps -T census of the 4 ranks, MPI_BIND='--bind-to core --map-by core', 2026-09-17 10:14:38
# loadavg 0.32 0.71 0.77 1/671 614620
pid 614624: 5 thread(s)  cores:   0  16  16  16  16 
pid 614625: 5 thread(s)  cores:   1   1  17   1   1 
pid 614626: 5 thread(s)  cores:  18   2   2   2   2 
pid 614627: 5 thread(s)  cores:   3  19  19  19  19 
# (a count of 0 means the run had already finished -- take more steps)
