NNSA awards contract for high-performance computers

Contract highlights efforts to integrate nuclear weapons complex: The Department of Energy’s National Nuclear Security Administration (NNSA) awarded Appro a contract to deliver the next generation of high performance computing systems to its three nuclear weapons laboratories – Lawrence Livermore National Laboratory, Los Alamos National Laboratory, and Sandia National Laboratories. “This is the first time NNSA has awarded a single contract for all three laboratories,” said Martin Schoenbauer, NNSA’s principal assistant deputy administrator for operations in the office of defense programs. “Combining the contract for each of NNSA’s three laboratories not only saves money, but continues to move NNSA towards a smaller more efficient nuclear weapons complex.” This $26.1 million contract is unique in that it will standardize computing systems among the three laboratories that make up NNSA’s Advanced Simulation and Computing (ASC) Program. Advanced simulations of nuclear weapons performance are a cornerstone of the Stockpile Stewardship Program, NNSA’s program to ensure the safety, security and reliability of the nation’s nuclear deterrent without underground testing A small business based in Milpitas, CA, Appro will provide new computer systems made up of modular “scalable units” that can be rapidly configured, lego style, into clusters of varying sizes and computing power. Each scalable unit represents about 20 trillion floating-point operations per second (teraflops) of computing power. By standardizing hardware, operating systems, and software, the three NNSA labs will improve efficiencies and cut the total cost of ownership by 30-50 percent. The new computers will provide the much needed “capacity” computing, running larger numbers of smaller jobs still requiring high performance machines. This allows NNSA’s larger supercomputers, or “capability” computer systems, to be dedicated to the larger, more complex calculations critical to stockpile stewardship. Previously, each laboratory had its own capacity computing machines. With different hardware and operating systems, these systems were not always compatible.