Running R on 32 cores for 2USD/h

May 14, 2014
By

(This article was first published on Sustainable Research » Renglish, and kindly contributed to R-bloggers)

Even though most modern computers already have four cores that can be used to speed up analysis – see here for a description. However, after starting to do more simulation studies it became apparent that this is highly impractical.

The solution: Running R via R-studio on Amazons cloud servers.

Thanks to Lois Aslett (http://www.louisaslett.com/RStudio_AMI/) this only takes four minutes toset up and connect to your own R-studio server in the cloud using the free tier program.

If you want to use this with more performant instances just follow the following steps.

1. Register with aws.amazon.com, including payment information etc.
2. Post a service limit increase. This can be done via http://aws.amazon.com/contact-us/ec2-request. You need an instance limit of at least 1. It may take amazon up to two business days to process this request…
3. Follow the description and links provided by Lois Aslett (http://www.louisaslett.com/RStudio_AMI/). These differ in location and thus pricing (right now the 32 core server costs between 1.7 and 2.3 USD per hour). When choosing and instance you can now select and launch more performant servers. I decided to go with “c3.8xlarge” which gives you 32 cores and at least for me adequate memory and storage.
4. Connect to your own 32-core R-Studio instance in the cloud.

To leave a comment for the author, please follow the link and comment on his blog: Sustainable Research » Renglish.

R-bloggers.com offers daily e-mail updates about R news and tutorials on topics such as: visualization (ggplot2, Boxplots, maps, animation), programming (RStudio, Sweave, LaTeX, SQL, Eclipse, git, hadoop, Web Scraping) statistics (regression, PCA, time series, trading) and more...



If you got this far, why not subscribe for updates from the site? Choose your flavor: e-mail, twitter, RSS, or facebook...

Comments are closed.