IT Infrastructure: IBM Supercomputers: 12 Ways HPC Is Leading to Scientific Breakthroughs
Earlier this year, IBM announced partnerships with Rice University and Rutgers University to advance supercomputing at their institutions and in their respective home states of Texas and New Jersey. The agreements became the latest of many partnerships IBM has with academic, scientific and research institutions that use IBM supercomputers. On March 30, IBM and Rice University announced a partnership to build the first IBM Blue Gene supercomputer in Texas. Rice announced a related collaboration agreement with the University of Sao Paulo in Brazil to share the administration and use of the Blue Gene supercomputer. IBM officials said Rice faculty will use the Blue Gene to further their own research and to collaborate with academic and industry partners on a broad range of science and engineering questions related to energy, geophysics, basic life sciences, cancer research, personalized medicine and more. Meanwhile, on March 27, Rutgers teamed with IBM to launch a high-performance computing (HPC) center at the university focused on the application of big data analytics in life sciences, finance and other industries. The center's primary goal is to improve the economic competitiveness of New Jersey's public and private research organizations. The HPC center will be part of the newly created Rutgers Discovery Informatics Institute, and will use supercomputing equipment and software provided by IBM in the project's first phase. Rutgers expects future expansion of the center to enable the university to assemble one of the world's most powerful academic supercomputers. Here, eWEEK looks at how 12 institutions are using IBM supercomputers.
The Argonne Leadership Computing Facility (ALCF) started deployment of Mira in January 2012 with the delivery of two single-rack systems. By early March, all of the Early Science Program (ESP) project application teams got their codes built and running on these racks. When completed this fall, Mira will have 48 racks and 786,432 processors, and weigh 104 tons. It will be 20 times faster and five times more energy-efficient than Argonne's current system, Intrepid, and capable of 10 quadrillion floating-point operations per second. The ALCF is committed to delivering 768 million core hours on Mira in 2013, based on a production schedule starting October 1, 2012, although it is likely that Mira will reach production status much sooner. When it goes into full production, more than 5 billion computing hours will be allotted to scientists on Mira every year. Any researcher in the world can apply for time on Mira to run programs for their experiments.