ACADEMIA

A new supercomputer for biomedical simulation and data analysis

The Computation Institute (CI) a joint initiative between The University of Chicago (UChicago) and Argonne National Laboratory (Argonne), is pleased to announce the introduction of Beagle, a 150 teraflops, 18,000-core Cray XE6 supercomputer that will support computation, simulation and data analysis for the biomedical research community. Made possible by a grant from the National Institutes of Health (NI) National Center for Research Resources (NCRR), Beagle will be housed in the new Theory and Computing Sciences (TCS) building at Argonne and will be available for use by UChicago researchers, their collaborators and other meritorious investigators nationwide.  “Innovative research requires access to the latest technologies," said NCRR Director Barbara Alving, M.D. "This high-performance tool will serve as a core resource that will help ensure scientists remain at the forefront of modern biomedical research.”

The system is named after the HMS Beagle, the ship that carried Charles Darwin on his famous scientific voyage in 1831. While Darwin’s Beagle enabled discoveries that established a unifying theory for all life sciences, the unique capabilities of UChicago’s Beagle will enable transformative innovation in basic, translational and clinical research leading to improved diagnostic strategies and life-sustaining medical treatment. Beagle is scheduled to dock in the TCS machine room by the end of the year, with her maiden voyage, for early adopters, set for the 202nd anniversary of Darwin’s birthday, February 12, 2011. The system should be placed into full production by the second quarter of the calendar year.

Ian Foster, director of the CI and principal investigator for the project, with UChicago’s team of technical and domain specialists, identified the need for a powerful computational environment that would serve the growing resource-intensive requirements of the biomedical research community. “Computation is fundamentally changing the nature of research in most disciplines, and biomedical researchers with access to advanced computational resources are more likely to make transformative progress,” said Foster. “We are excited about the opportunity to advance the biomedicine research frontier and greatly appreciate the support of the NIH-NCRR for this initiative.”

Twelve NIH-funded biomedical research teams, with more than 100 faculty, students, staff, and postdoctoral scholars from multiple research institutions, will be among Beagle’s early adopters. Teams presented a case for additional and dedicated computing resources in support of research spanning multiple arenas, including the prevention and treatment of cancer, improved management of burn victims, better drug design, genetics and inherited disorders, and the development of patient-specific medicine. The vast diversity among the teams’ research strategies will provide computational scientists, high-performance computing (HPC) consultants and systems administrators with a rich proving ground so that Beagle can be finely tuned for a range of biomedical research applications before going into full production.

“The convergence of whole-genome experimental data sets and digitalized clinical data from hospital patient records is changing the landscape of biomedical research,” said Conrad Gilliam, dean for research and graduate education in the UChicago biological sciences division. “UChicago is poised to lead this exciting new phase of discovery with nationally recognized leadership in high-end computation, evolutionary genetics and genomics, and translational research, along with a history of interdisciplinary problem-solving. The arrival of Beagle will challenge our best minds as we forge new modes of inquiry to extract meaningful biological and medical information from these massive tomes of data,” he added.

UChicago’s Benoit Roux, with collaborators from Cornell University and the University of Maryland, and the additional researchers who supported this project, will be an early Beagle user. Roux, director of the Computational Core Facility of the Membrane Protein Structural Dynamics Consortium, which was recently funded by NIH, anticipates that Beagle will be an important resource for advancing scientists’ understanding of several membrane proteins that play a critical role in human health because their malfunction is often the cause of disease. "To understand and comprehend how membrane proteins work, one must be able to visualize the manner in which these molecular machines move and change their shape, atom by atom, as they go about their business. Beagle will enable us to perform the detailed computations necessary to reach this goal," says Roux.

Beagle is anticipated to place among the top 50 fastest supercomputers in the world and will be one of the fastest systems fully devoted to life sciences. Beagle uses a unique combination of AMD multi-core processors, Cray’s powerful Gemini system interconnect, and 3D torus topology in an infrastructure designed to scale to more than 1 million processor cores. Following a thorough assessment of five different vendors’ HPC solutions and two hybrid systems, Foster’s team recognized Cray’s reputation for scalability, reliability, flexibility and efficiency. Designed for scalability, with future investments, Beagle can be upgraded to achieve sustained petaflops performance.  

“We are honored to provide UChicago and NIH with one of the world’s most powerful life sciences supercomputers,” said Barry Bolding, vice president of Cray’s products division. “Cray specifically designs systems to improve scientific productivity, and we look forward to the important health-related research results that these innovative teams will produce with the aid of the Cray XE6 supercomputer.”