Microsoft Machine Learning R Server

Microsoft R Server is a standalone server version of the R engine, built on top of the R open source platform but modified to suit the needs of the enterprise environment and heavy usage of analytical purposes, from statistical analysis and data mining to machine learning and any big data tasks.

It is fully compatible with CRAN/MRAN/GitHub repositories and any R legacy code. The ScaleR functions and algorithms available in the RevoScaleR package are capable of parallel and multi-threated data processing with no limitations known to R Open or R Client. Computations are prepared for large datasets—in most cases, datasets that easily exceed the server memory size. This can be done due to a special external data frame format that enables multi-threaded computations and enables R-based applications that can perform parallel computations and deployable solutions across multi-platforms with the help of ConnectR and DeployR. Disk scalability is also available with Microsoft R Server.