IBM and the Department of Energy’s Lawrence Livermore National Laboratory (LLNL) have announced a “broadening” of their two-decade partnership, one that could give U.S. companies a supercomputing boost over global competition. Under the terms of the expanded partnership, computer and domain science experts from IBM Research and LLNL will collaborate with representatives from various industries to figure out how high-performance computing (HPC) can be applied to developing new technologies in a number of fields, including biology, applied energy, manufacturing, data management and informatics. The collaboration, known as Deep Computing Solutions, will take place in the LLNL’s High Performance Computing Innovation Center (HPCIC), which was created expressly for this purpose. By summer, the parties involved plan on positioning a 24-rack IBM Blue Gene/Q system (called “Vulcan”) to deliver five petaflops’ worth of processing power to the effort. One petaflop is the equivalent of one quadrillion floating-point operations per second. “Maintaining a technological edge over the competition in the global marketplace is vital to both national security and the country’s economic prosperity,” Frederick Streitz, director of the HPCIC, wrote in a June 27 statement. “Deep Computing Solutions will be an important ingredient of the HPC Innovation Center, building on both IBM and LLNL’s mutual experience in applying HPC to complex technical problems.” According to data released by IBM, more than 70 percent of members of the National Center for Manufacturing Sciences (NCMS) believed that increased adoption of advanced computing would translate into competitive advantages within the marketplace. However, only 6 percent of small-to-midsize manufacturers in the United States actually take advantage of HPC resources. The use of HPC can translate into significant monetary savings for companies. Boeing, for example, is using all that massive computing power to redesign commercial jets’ vertical tails in ways that could save $300 million in annual fuel costs. LLNL, based in Livermore, Calif., is already home to the world’s fastest supercomputer, another IBM BlueGene/Q system known as “Sequoia.” That massive machine can utilize its 1,572,864 processing cores to deliver 16.32 sustained petaflops on the Linpack benchmark, a widely accepted performance yardstick based on the ability to solve a dense system of linear equations. Sequoia recently topped the influential Top500 list of the world’s fastest supercomputers. Researchers at the National Nuclear Security Administration (NNSA) use it conduct virtual nuclear weapons tests, as a way to monitor the nation’s aging stockpile without actually setting off a bomb.   Image: IBM