I am trying to understand if basic matrix algebra operations in R, e.g. A %*% B, are inherently multithreaded, such that when I run the same script on a server with 72 cores, all matrix algebra computations will take advantage of the 72 cores.
Apparently, R on Linux/MacOS is inherently single-threaded (SO question).
But matrix algebra in R uses ATLAS, based on BLAS, which apparently can take advantage of multiple cores if built as such.
How do I check if ATLAS used by my R installation is built with mulithreading? I downloaded R 3.5.3 binary for MacOS from CRAN.
The R FAQ does not mention if Rblas is built multithreaded or not.
Furthermore, how do I control the number of cores used for matrix algebra in R
