Short update to Speed up R by using a different BLAS implementation/:
- MKL is overall the fastest
- OpenBLAS is faster than its parent GotoBLAS and comes close to MKL
A = matrix(rnorm(n*n),n,n) A %*% A solve(A) svd(A)
Short update to Speed up R by using a different BLAS implementation/:
A = matrix(rnorm(n*n),n,n) A %*% A solve(A) svd(A)
May I ask how you compiled the Rblas.dll based on OpenBLAS for the above comparison? I have been having serious difficulty getting OpenBLAS to compile into a working Rblas.dll for 64bit Windows that passes all the checks.
I did not compile OpenBLAS myself for this. I used a binary (DLL) from the internet and renamed Rblas.dll. I uploaded it here: https://www.dropbox.com/sh/ooo044s2y5ev7xz/o9bwnuMZ6d/rblas_openblas.zip