Driving Discovery and Innovation with Community Clusters

BrandPost By Chris A. Jones
Jan 16, 2020
AnalyticsBig DataHadoop

Purdue University arms its diverse researcher communities with a suite of leading-edge high performance computing clusters for compute- and data-intensive investigations.

browncluster 1200
Credit: Dell Technologies

Since its founding in 1869 as a college dedicated to teaching agriculture and “the mechanic arts,” Purdue University has served as a catalyst for scientific research that advances the public good. Today, Purdue remains true to its roots by supporting research programs that drive advances in fields ranging from engineering and artificial intelligence to sustainability and health.

To enable these world-class research programs, Purdue and its faculty researchers invest in leading-edge high performance computing systems that drive compute- and data-intensive scientific investigations. And this is where the Information Technology at Purdue (ITaP) Research Computing organization enters the picture.

ITaP Research Computing provides advanced computational resources and services to support Purdue

faculty, staff and student researchers. As part of this mission, the organization operates a suite of top-ranked cooperative supercomputing clusters. In any given year, Purdue operates up to five flagship community cluster systems, which are integrated into one large supercomputer with the right technologies to solve some of the biggest problems in science, engineering, physical sciences and life sciences.

Currently, two of these systems are based on Dell EMC solutions for HPC and AI — the clusters known as Brown and Gilbreth.

  • The Brown cluster, built through a partnership with Dell EMC and Intel and deployed in 2017, is optimized for researchers running traditional, tightly coupled science and engineering applications.
  • The Gilbreth cluster, in service in the first half of 2019, is optimized for machine learning, artificial intelligence and other accelerator-fueled applications.

Both clusters are built with versatile Dell EMC PowerEdge™ servers with Intel® Xeon® Scalable processors. They are designed to allow users to manage large, complex datasets, gain insights quickly, accelerate product innovation and drive scientific explorations that weren’t previously possible. Lustre storage, InfiniBand and Ethernet networking are also part of the solution.

Researchers across the Purdue community put Brown, Gilbreth and the other community clusters to work to drive discovery in virtually all academic disciplines. All told, Purdue has about 1,300 researchers using the HPC clusters, including graduate students, undergraduate students and faculty members.

With the community cluster model in use at Purdue, which builds a new HPC/AI system every year, these researchers know they will always have access to a cutting-edge supercomputer to accelerate discovery and innovation.

To learn more