We recently updated the Oracle R Distribution (ORD) benchmarks for version 3.3.0. ORD is based on open source R-3.3.0 and adds support for dynamically loading linear algebra performance libraries installed on your system. This includes Intel's Math Kernel Library (MKL), AMD's ACML, and Sun Performance Library for Solaris, which enable optimized, multi-threaded math routines to provide relevant R functions maximum performance.
The benchmark results demonstrate the performance of Oracle R Distribution 3.3.0 with and without dynamically loaded MKL. We executed the community-based R-Benchmark-25 script, which consists of a set of tests that benefit from faster matrix computations. The tests were run on a 24 core Linux system with 3.07 GHz per CPU and 47 GB RAM.
On average, Oracle R Distribution with dynamically loaded MKL and 8 threads is 50 times faster than Oracle R Distribution (and open source R) using R's internal BLAS library (Netlib) with 1 thread. Matrix multiplication is an astounding 94 times faster than single-threaded R, and principal components analysis is 32 times faster.
As always, ORD is free to download, use and share, and is available from Oracle's Open Source Software portal. Oracle R Distribution 3.3.0 will be supported with the upcoming version of Oracle R Enterprise 1.5.1. Installation Instructions are in the Oracle R Enterprise Installation and Administration Guide. Look for a future blog post announcing it's release.