Sign In

Communications of the ACM

ACM TechNews

High Throughput Computing Helps Ligo Confirm Einstein's Last ­nproven Theory

The Laser Interferometer Gravitational-Wave Observatory facility in Livingston, LA.

The Laser Interferometer Gravitational-Wave Observatory is using software pioneered at the University of Wisconsin-Madison to help researchers detect gravitational waves created by the collision of two black holes 1.3 billion years ago.


Since 2004, the HTCondor High-Throughput Computing (HTC) software pioneered at the University of Wisconsin-Madison (UW-Madison) has been analyzing data generated by the Laser Interferometer Gravitational-Wave Observatory (LIGO). LIGO's latest use for the software is helping researchers detect gravitational waves caused 1.3 billion years ago by the collision of two vast black holes.

HTCondor taps the power of tens of thousands of networked computers to run a large series of computational tasks via advanced HTC technologies. Former UW-Madison researcher Peter Couvares says the software is a major advantage for big-science projects because "we know it will work--that's the killer feature of HTCondor."

He notes HTCondor works because built into its core software is the assumption it is impossible to assume a network of thousands of individual computers will not experience local failures. "The biggest trick of all the data analysis in LIGO is to come up with a better signal-to-noise ratio," Couvares says.

University of Southern California Information Sciences Institute professor Ewa Deelman says HTCondor, in conjunction with her Pegasus workflow automation engine, helps remove technological obstacles for scientists. "I think the automation and the reliability provided by Pegasus and HTCondor are key to enabling scientists to focus on their science, rather than the details of the underlying cyberinfrastructure and its inevitable failures," she says.

From UW-Madison News
View Full Article


Abstracts Copyright © 2016 Information Inc., Bethesda, Maryland, USA


No entries found

Sign In for Full Access
» Forgot Password? » Create an ACM Web Account