SCIENCE

The University of Minnesota Supercomputing Institute for Advanced Computational Research (MSI) and SGI has announced that MSI will deploy a new high performance computing (HPC) system this fall, featuring an SGI Altix UV 1000 supercomputer, to accelerate its research program. This new system, named Koronis, is made possible by a National Institutes of Health (NIH) grant awarded to the University of Minnesota in June 2010.

Assembling a team of more than 30 University of Minnesota faculty colleagues in the life sciences community with similar specialized compute needs, including four major user groups in multi-scale modeling, chemical dynamics, bioinformatics and computational biology, and biomedical imaging, Professor Darrin York from the Department of Chemistry, with the assistance of HPC specialists at both SGI and MSI, submitted the grant proposal to the NIH in May 2009. The proposal was for a complete High End Computing (HEC) solution that would accommodate the needs of a broad spectrum of NIH-supported researchers. This HEC solution will facilitate breakthroughs in biomedical research and significantly impact human health.

MSI named the system Koronis after a Minnesota lake; it will feature a 1,152-core SGI Altix UV 1000 system with a shared-memory architecture in which each core can access all 3.1 terabytes of memory directly. Koronis will also include high-performance visualization workstations to handle large-size biomedical data and high-performance and -fidelity data storage to ensure adequate data analysis and processing.

These high-performance visualization workstations will enable researchers to develop next-generation sequencing, which can help provide a more complete picture of genomes, producing thousands of gigabytes of data. With next-generation sequencing, along with other large-sized data, it is hoped that Koronis will facilitate breakthroughs in biomedical research.

MSI Principal Investigators (PIs) are already planning the jobs they will run on Koronis. Professor York will develop new theoretical tools for biocatalysis and multi-scale quantum models for Ribozyme Catalysis, using computational methods to simulate complex chemical reactions that are catalyzed by biological molecules and, in particular, RNA. These molecular simulations, using so-called "multi-scale quantum models," provide potentially powerful tools for studying catalytic mechanisms as well as detailed characterization of the transition state ensemble that are difficult, if not currently impossible, to directly observe by experiment. The simulations involve tens to hundreds of thousands of atoms, and require sampling tens to hundreds of millions of configurations along the reaction path. To carry out these large-memory, data-intensive computational simulations, MSI required a supercomputer that can handle an enormous capacity of data, which they found in SGI's Altix UV 1000.

Professor Jiali Gao, recent recipient of the IBM Faculty Award, has been using MSI supercomputers for many years, the last few to support his research in density functional theory (DFT). Koronis will enable Gao to investigate faster and more efficiently DFT, alternative reaction pathways and the effects of amino acid mutations. As well, the research groups of Donald Truhlar, George Karypis, David Largaespada, Kelvin Lim and Bin He will also be continuing their work on the new supercomputer.

Replacing the recently retired SGI Altix computer system, Koronis will join the other MSI supercomputing resources housed in Walter Library. These other systems, Itasca, Calhoun, Blade and Elmo, are available to all researchers at institutions of higher education in Minnesota. More than 4,000 active users across a wide range of disciplines utilize MSI's diverse computational resources, making MSI a focal point of collaborative research at the University of Minnesota. MSI currently supports more than 400 active research groups by providing complete HPC environments, including systems, software, storage support and services. MSI resources have helped these groups obtain almost $150 million in external funding in the past academic year.