Feature - Introducing Condor at RIT
Grid computing at the Rochester Institute of Technology, New York, is not a new initiative, but with the advancement of technology and recent additions to the resource pool, RIT is setting its sights on further expanding its grid presence and contributions.
The Condor computing infrastructure harnesses the power of idle computers across the RIT campus, and has been in use sporadically since 1999. It has recently gained a campus-wide foothold thanks to a pushstart from Research Computing at RIT.
The Condor network has expanded to over 1000 processors among three flocks, with a goal of reaching 2000 processors in the near future.
This computing network spans a multitude of operating systems and machine architectures, with installations on dedicated clusters, personal workstations and in a large number of computer labs on campus.
New horizons for research
Utilizing the newly harnessed computing power, researchers at RIT have begun to broaden the scope of their inquiries. The future of Condor at RIT also seems bright, with many other potential projects that leverage the Condor resources in development.
Andrew Darling, an RIT bioinformatics graduate student, is studying the evolution of reptiles with supervisor Larry Buckley.
“The software that compares the DNA sequence of the around 50 species takes 30 CPU days to run, since it compares nearly every evolutionary path,” Darling says. “By splitting the task across the 1000-processor Condor flock, the same task should take about 45 minutes, in theory.
Growing new connections
RIT is not simply content to service internal research; we are undertaking a variety of projects aimed towards collaboration with statewide and worldwide computing projects.
RIT is currently participating in the New York State Grid, joining 20 other participating institutes in the state. NYSGrid is a virtual organization of the Open Science Grid, another organization with which RIT is working closely.
We are very excited to have grown our Condor flock so dramatically in such a short time and are anxious to share these new capabilities with researchers around the world.
- Brent Strong, Research Computing, RIT