Parallel BLAS/LaPack

I have a large regression problem, roughly 200,000 observations on 20,000 variables. Getting an estimate can take days. The problem is running on a Windows 10 machine with 26 cores.

I have been reading that it is possible to install a multi-threaded version of the underlying BLAS library which might give an order of magnitude speed-up. Can I get advice on this? (Preferably using very small words that are easy to understand :slight_smile:)


Let me add a little to my own question . On further reading this seems to mean switching to the Microsoft version of R. Is that a reasonable thing to do?

It's not an unreasonable thing to do.

1 Like

Switch to Microsoft R Open.

This topic was automatically closed 21 days after the last reply. New replies are no longer allowed.

If you have a query related to it or one of the replies, start a new topic and refer back with a link.