Setting the parameter cores to NULL, or 8, or 4 ... on a Mac2 Pro running in R I never see more than 25% CPU usage.
I understand on Windows the overhead defeated performance gains; is it the same on Mac?
I've 12 cores and 16gb in theory sitting fairly idle. Is the solution a simple fix on my end?
I've tried adding this to the beginning of my script, but no difference has ensued:
Sys.setenv(R_FUTURE_FORK_ENABLE = "true")
options(future.fork.enable = TRUE)
R version 4.5.2
Robyn 3.12.1
Nevergrad installed via conda (python 3.13 instance created)
Setting the parameter cores to NULL, or 8, or 4 ... on a Mac2 Pro running in R I never see more than 25% CPU usage.
I understand on Windows the overhead defeated performance gains; is it the same on Mac?
I've 12 cores and 16gb in theory sitting fairly idle. Is the solution a simple fix on my end?
I've tried adding this to the beginning of my script, but no difference has ensued:
Sys.setenv(R_FUTURE_FORK_ENABLE = "true")
options(future.fork.enable = TRUE)
R version 4.5.2
Robyn 3.12.1
Nevergrad installed via conda (python 3.13 instance created)