Microsoft quietly unwraps a big-data analytics platform based on R

4:55 PM

Microsoft quietly unwraps a big-data analytics platform based on R

whatthingnews.com : Though commonly used for statistics, R 'looks like Cobol to sophisticated developers,' one analyst says
Microsoft this week took the next step following its acquisition of Revolution Analytics last year by unwrapping an advanced-analytics platform of its own based on the open source R programming language. 

Dubbed Microsoft R Server, the platform works with Windows, Linux, Hadoop and Teradata software. It's now available on the Microsoft Developer Network and via Microsoft's DreamSpark educational program. 
Last April Microsoft completed its acquisition of Revolution Analytics, provider of the Revolution R Enterprise (RRE) product on which R Server is now apparently based. Though Microsoft has renamed the software R Server in the Linux, Hadoop and Teradata versions, it's retaining the RRE name in the Windows version of the software. 

Microsoft R Open, which the company calls an "enhanced distribution of R from Microsoft," is said to be a prerequisite for both R Server 2016 and RRE for Windows 2016. 





Microsoft declined to comment with further detail.

In September, RRE made its debut on Azure; the following month, R support popped up in a preview of Microsoft SQL Server 2016. 

The R Server platform supports a variety of big-data statistics, predictive modeling and machine-learning capabilities, all compatible with the R programming language, wrote Lee Stott, a technical evangelist for Microsoft UK, in a blog post on Wednesday. 

Users can run R scripts in a high-performance, parallel architecture to explore, model and predict at scale, Stott added. 

The R language has become increasingly popular for use in statistics and analytics. By acquiring Revolution, Microsoft sought to shore up its own advanced-analytics capabilities, said Nik Rouda, a senior analyst with Enterprise Strategy Group. 

It's "not trivial" to perform advanced analytics in SQL, for example, so Microsoft "wanted R to be their entry point," Rouda said. 

Now, its release of R Server is the next logical step in executing on that plan.
Microsoft's Machine Learning technology can use R "behind the scenes" for data preparation, statistics and machine learning, noted Mike Gualtieri, a principal analyst with Forrester Research. 

"It's a good move for Microsoft, but I certainly hope they don't put all their data-science 'eggs' in the R basket," Gualtieri added. 

"I know R is popular among data scientists, but I think most computer scientists would think that it is not a very elegant, efficient or performant programming language for any purpose," he said. "R looks like Cobol to sophisticated developers." 

Instead, Python and other more modern programming languages are increasingly coming into favor, he said. 

Microsoft rolls out its R Server big-data analytics line-up

whatthingnews.com : Microsoft R Servers for Hadoop, Teradata and Linux -- seemingly, the renamed version of Revolution Analytics' commercial versions of its analytics products -- are available to MSDN subscribers.

Microsoft quietly made available this week a new family of R Server data-analytics products.
Thanks to a tweet from Tero Alhonen, I found listings on MSDN for a family of new Microsoft R Server products with January 4, 2016 release dates.
The new line-up includes:
  • Microsoft R Server for Hadoop on Red Hat
  • Microsoft R Server for Teradata DB
  • Microsoft R Server for Red Hat Linux
  • Microsoft R Server for SuSE Linux
  • RRE, which is Revolution R Enterprise, I believe. (Thanks, Marko!)
Here's the description on MSDN as to what these R servers are:
"Microsoft R Server is a fast and cost-effective enterprise-class big data advanced analytics platform supporting a variety of big data statistics, predictive modeling and machine learning capabilities. Microsoft R Server includes Open Source R and is fully compatible with R scripts, functions and CRAN packages, and offers a variety of analytics capabilities including exploratory data analysis, model building and model deployment."
R is a programming language for computational statistics. Microsoft last April completed its acquisition of Revolution Analytics, the maker of a distribution of the R programming language for statistical computing and predictive analytics, for an undisclosed amount.

Microsoft previously announced it plans to build R support directly into SQL Server 2016. Microsoft also announced last year plans to integrate Revolution's R distribution into the Azure HDInsight and Azure Machine Learning services.

Microsoft execs also said last year they would continue to support the open-source evolution of R, as well as the commercial distributions of Revolution R across multiple platforms, including Linux, Teradata and Hadoop.

I believe the new R Server products available to MSDN subscribers are the renamed versions of the promised commercial distributions of Revolution R. I've asked Microsoft officials for clarification, pricing and more details. No word back so far.




Update (January 6): There's a new Microsoft blog post that notes the R Servers are now also available to academics and students via DreamSpark. And there's a page with information about "Microsoft R Open," which is described as the "enhanced distribution of R from Microsoft." Microsoft R Open is a prerequisite for Microsoft R Server 2016 and RRE 2016, according to that page.

(Thanks, again, to Tero Alhonen for these latest links to the R Server information.)


Previous
Next Post »