Availability gives cloud the edge over dedicated machines

Availability gives cloud the edge over dedicated machines

Researchers in a National Institutes of Health lab were wondering if cloud computing offered any advantages over their own high-performance cluster designed for biosciences. To find out, they did what scientists usually do -- they ran an experiment.

NIH’s National Heart, Lung and Blood Institute (NHLBI) began to test the use of third-party cloud providers approved by the Federal Risk and Authorization Management Program. Although the cloud was ideal for hosting the institute’s website and cold storage, the institute wanted to see if it provided advantages for compute-intensive research.

The institute ran a test that pitted a third-party cloud against Biowulf, a 20,000+ processor Linux cluster in NIH’s high-performance computing facility, and recruited Maria Mills, a postdoc in NHLBI’s Laboratory of Single Molecule Biophysics, to compare the platforms. Mills studies how a protein interacts with a small molecule. According to Mills, the lab runs simulations that are computationally expensive, as they require quickly and repetitively calculating a massive amount of data to see how every part of the protein interacts with all parts of the molecule.

Though the performance on Biowulf and the cloud was similar, Mills said, using the cloud saved time overall, because she didn’t have to wait for available computing time on the shared Biowulf system. Working with the cloud provider’s technicians, she was able to design a virtual computer structure and write scripts that not only optimized the process, but also allowed Mills to take advantage of temporarily dedicated compute resources.  

Using the cloud for her research meant Mills did not have to wait for her turn on Biowulf or “worry about other people not having access to resources they needed,” she told the NIH Record. “It was a virtual computer cluster that I built and I’m using, and when I’m done with it I just shut if off.

About the Author

Amanda Ziadeh is a Reporter/Producer for GCN.

Prior to joining 1105 Media, Ziadeh was a contributing journalist for USA Today Travel's Experience Food and Wine site. She's also held a communications assistant position with the University of Maryland Office of the Comptroller, and has reported for the American Journalism Review, Capitol File Magazine and DC Magazine.

Ziadeh is a graduate of the University of Maryland where her emphasis was multimedia journalism and French studies.

Click here for previous articles by Ms. Ziadeh or connect with her on Twitter: @aziadeh610.


Reader Comments

Thu, Sep 15, 2016 Todd Crofton MD

Interestung but could you provide info about the CSP, the performance similarities and the configuration and setup they used to create a virtualized parallel processing computing cluster and the programming language used to create test scenarios in the virtual environment. It seems the article is lacking substance. Would like to know a little more as it relates to details.

Please post your comments here. Comments are moderated, so they may not appear immediately after submitting. We will not post comments that we consider abusive or off-topic.

Please type the letters/numbers you see above