Faculty meeting to discuss new all-GPU research supercomputer set for Oct. 16

I'm interested in attending

October 16, 2018  1:30pm
LWSN 1142

A luncheon for faculty, staff and graduate students will be held from noon to 1:30 p.m. on Tuesday, Oct. 16, in the Lawson Computer Science Building, Room 1142, to discuss the specifics of the new all-GPU community cluster research supercomputer to be built this fall.

An update on Purdue’s research computing resources will be given by ITaP Research Computing staff members, followed by an overview of the proposed new supercomputer and a question and answer session. Food and drink will be provided. Register here.

The new cluster, which will be known as Gilbreth, will be optimized for machine learning, artificial intelligence and other graphical processing unit-enabled applications. With Gilbreth, ITaP Research Computing is building the system with the goal to provide a cost-effective way to deliver as many GPUs as possible to the growing community of faculty that need them for their research, says Preston Smith, ITaP director of research services and support.

Gilbreth will be the tenth research computing system built through the Community Cluster Program. There are now nearly 200 faculty partners from all of Purdue’s primary colleges and schools using the community clusters and other services operated by ITaP for research, spanning more than 35 science, engineering, life sciences and social science disciplines.

ITaP Research Computing installs, administers and maintains the community clusters, including security, software installation and user support, so researchers can concentrate on doing research rather than on operating a high-performance computing system.

Centralized, large group purchasing allows Purdue to negotiate better prices with vendors and to buy more computing capacity than individuals or campus units could otherwise afford.

Community clustering also maximizes use by sharing computing power among the faculty partners whenever it is idle. Researchers always have ready access to the cluster capacity they purchase, but they also can share capacity fellow researchers aren’t using at the moment. This offers users access to substantially more computational power if needed and keeps the machines busy.

Additionally, the community clusters provide high-end computational resources coupled with a variety of research data storage, moving and sharing services including the Research Data Depot and the Fortress HPSS Archive.

To learn more about the Community Cluster Program, contact Smith, psmith@purdue.edu or 49-49729.

Originally posted: October 9, 2018  3:16pm