R is fast becoming a powerful tool for high-performance computing: the art making computational problems that take a long time to process run faster through the use of multiprocessor computers or computer clusters.
Dirk Eddelbuettel has helpfully created a list of R resources for high-performance computing as a new task view on CRAN. On that page you'll find lists of packages for explicit and implicit parallel computing, grid computing, and out-of-memory data analysis for large data sets. You'll also find links to tools for batch scheduling and resource managers, and utilities to help with compilation and code profiling.
The list includes the NetworkSpaces (nws) package, which is the basis of the ParallelR module included with REvolution R Enterprise. High-performance computing is a particular focus of the development team here at REvolution, and as a result REvolution R has significant advantages on multiprocessor machines and clusters.
If you're interested in developments in high-performance computing with R, you might want to join the R-sig-hpc mailing list, which is devoted to discussion of the topic (check the archives for previous discussions). There's also a list of high-performance computing resources on the REvolution Computing website.
Comments