R on Xeon Phi

[CPati2]

Hi All,

Is it possible to make use of OpenMP thread distribution and affinity for R on Xeon Phi? If yes, please share resources that can help understand how to optimize R for Xeon Phi.

Since, R code runs on R engine and doesn't go through the linking process, I am not sure how this will work out.

Any suggestion will help.

Thanks.

[Sebastian_S_Intel · Accepted solution]

Hi,
To take advantage of OpenMP you have to use library that uses it like MKL.
Here are interesting resources which may be helpful:

• - https://software.intel.com/en-us/articles/build-r-301-with-intel-c-compiler-and-intel-mkl-on-linux
• - https://software.intel.com/en-us/articles/using-intel-mkl-with-r

[jimdempseyatthecove]

Google:   multithreaded r programming

http://www.cs.toronto.edu/~radford/ftp/pqR-guelph.pdf
http://www.win-vector.com/blog/2016/01/parallel-computing-in-r/
http://garthtarr.com/parallel-computation-in-r/

There are additional links that may be of interest via the above search

Jim Dempsey