R Statistical Language Clustering With Windows HPC Server 2008

by Sam Dean - Nov. 17, 2008Comments (0)

Earlier this year, we covered R, a surprisingly robust open source statistical language. Revolution Computing has just announced that it is integrating its RPro commercially supported distribution of R into Microsoft's Windows HPC Server 2008, which is focused on clusters. A module within RPro, called ParallelR, enables cluster computing. Will this make high-performance computers and applications easier to deploy?

"Windows is the dominant choice of R Language users on workstations. With the release of Windows HPC Server 2008, we expect the supercomputer community to embrace this new option" said Revolution Computing CEO Richard Schultz.

R is used by statisticians and scientists to derive answers to problems that usually involve very large data sets, and is used widely for predictive models. Revolution Computing will be demonstrating RPro and the ParallelR module in Microsoft's booth at the Super Computing 2008 conference.

Many people consider clusters easier to build and deploy on Linux than on Windows, but Microsoft has been increasingly focused on high-end computing. Although Microsoft will be tapping a commercial version of R in this case, R has a firmly rooted open source history and community. Given the positive reactions R gets from most people with the requisite statistical knowledge, I bet some people within the high-performance computing community will benefit from this hookup.