Along with the integrated R services on SQL Server 2016, Microsoft is also offering enterprise level Microsoft R Server as a standalone installation. Some of the documentation on MSDN shows that the standalone MS R Server is different from a data science Microsoft R Client.

This is a picture from the MSDN page (bottom left): https://msdn.microsoft.com/en-us/library/mt696069.aspx
R Server vs R Client on MSDN Page

The data science client link takes you here: https://msdn.microsoft.com/en-us/library/mt696067.aspx

But the set-up wizard comments as if they're both the same thing.

enter image description here If anyone has any clarity on this, please let know.


1 Answer 1


Looks like I found an answer to my own question on digging further.

While Microsoft R standalone Server addresses the in-memory limitations of open source R by adding parallel and chunked processing of data, Microsoft R Client is in-memory bound; i.e. it can only process datasets that fit into the available local memory. However, it can operate on large volumes of data if the compute context is set to an instance of Microsoft R Server like the SQL Server R Services or R Server for Hadoop etc.

Microsoft R Client has almost all of the architectural features as a MS R Server - DistributedR, ScaleR (RevoScaleR package), ConnectR. DeployR is unavailable in the MS R Client.

Microsoft R Client also has the Math Kernel Library for the non-scaleR functions, and can use up to 2 threads for ScaleR functions in a local compute context.

One advantage that the MS R Client has on the MS R Server is that it is free for everyone. MS R Server (standalone) requires a commercial license.

More details here: The Microsoft R: A Product Comparison

  • One more thing to add is that you can't install the R portion from the SQL 2016 media alone on older OS (like Windows 7). You can install the client from download though. Jun 29, 2016 at 23:24
  • You could try SQL R for free with SQL Server Developer edition--which includes all the Enterprise features. "With Enterprise Edition, you also get the ScaleR libraries to overcome R’s inherent performance and scale limitations."
    – Sting
    Oct 12, 2016 at 15:25

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.