New community cluster Bell arriving this year, space available for purchase now
Purdue’s newest community cluster supercomputer will be twice as fast as the retired Conte cluster, which was the nation’s fastest campus supercomputer at the time it was built just seven years ago.
The new cluster, known as “Bell” after Purdue nursing professor and minority advocate Clara Bell Sessions, is optimized for traditional, tightly-coupled science and engineering applications. Bell will consist of 448 Dell compute nodes featuring two 64-core AMD Epyc “Rome” processors with 256 GB of memory. The cluster will also include 8 large memory nodes, each with 1 TB of RAM, and a GPU subsystem with 16 AMD M150 GPUs. The system will feature a 5 PB Lustre parallel filesystem, with its top performance 50% faster than the storage powering the Brown community cluster.
Bell will be further enhanced by a composable container subsystem that will enable new types of science to be served by the community clusters. This subsystem will support a wide range of non-batch technologies, such as databases, gateways, middleware, and web services; and provide new types of storage interfaces beyond the parallel filesystem.
Purdue researchers can now buy capacity in Bell through ITaP Research Computing’s cluster orders website. The cluster will be available for early users by fall 2020, and will have a six year lifespan. Bell delivers the lowest cost per FLOP and per core of any community cluster system to date.
Bell’s namesake, Clara Bell Sessions, was a nursing professor and the director of continuing education in the School of Nursing, who helped establish Purdue’s Minority Student Nurses Association, now known as the Diversity in Nursing Association, and the Minority Faculty Fellows programs.
After her death in 1996, Purdue’s Black Caucus of Faculty and Staff created the Clara E. Bell Academic Achievement Award for the senior in nursing or health sciences with the highest grade point average. In 2013, she was posthumously awarded Purdue’s Title IX Distinguished Service Award for her contributions to the advancement of gender equity in education.
More information about Clara Bell Sessions is available in an ITaP biography.
Bell is the latest addition to Purdue’s community cluster program, which began in 2004 with just four faculty partners. Today the program has more than 200 active faculty partners from all three Purdue campuses, all of Purdue’s primary colleges and schools and 60 different departments. In 2019, ITaP Research Computing delivered more than 350 million computational hours to community cluster users.
Community clustering makes more computing power available for Purdue researchers than faculty and campus units could individually afford, and provides this crucial service at the lowest cost to the institution. ITaP Research Computing installs, administers and maintains the community clusters, providing reliability, security, software installation and expert user support.
“Many faculty members in EAPS do highly computational research, whether simulating severe weather, earthquakes, or asteroid impacts or analyzing massive geoscientific datasets, so having world class high-performing computing on campus is incredibly important for us,” says Matthew Huber, professor of earth, atmospheric and planetary sciences. “The cluster computing model, because of its stability and support, gives our faculty the ability to carry out world-class, long-term research plans knowing that the clusters will grow to meet our computing needs.”
Gerhard Klimeck, a professor of electrical and computer engineering and director of Purdue’s Network for Computational Nanotechnology, uses the community clusters for his research in nanoelectronics. He says the community cluster program gives faculty a lot of bang for their buck, since there’s no need to pay for hardware or invest graduate student time in maintaining a computer, and users share capacity other researchers aren’t using. “As computational demands of an individual group ebb and flow, researchers benefit from available free resources when others are not running to their full ownership extent,” he says.
To learn more about the Bell community cluster or other Research Computing resources, contact Preston Smith, executive director of research computing, firstname.lastname@example.org or 49-49729. An informational webinar on the Bell cluster will be hosted on July 31. Please register to attend here: RCAC 2020 Cluster Town Hall