Dear, I am not sure to understand if it is expected or not.
Let install R, guix package -i r, and then run the small script: > N <- 9000;A <- array(0.1, dim=c(N,N));B <- array(0.1, dim=c(N,N)) > dim(B %*% A) Monitoring with htop, only 8 of my 16 cores are used. Why ? And if I have more, then again only 8 cores will be used. Is it from OpenBlas? I miss the compilation option forcing this behaviour. I expect that R will use as many core as available. Right? Thank you for any insight. All the best, simon