BlueGene/L was a revolutionary, low-cost machine delivering extraordinary computing power for the nation's Stockpile Stewardship Program. It earned the number 1 position on the TOP500 list from November 2004 to June 2008.
Located in the Terascale Simulation Facility at Lawrence Livermore National Laboratory, BlueGene/L was used by scientists at Livermore, Los Alamos, and Sandia National Laboratories. The 596-teraflops machine handled many challenging scientific simulations, including ab initio molecular dynamics; three-dimensional (3D) dislocation dynamics; and turbulence, shock, and instability phenomena in hydrodynamics. It was also a computational science research machine for evaluating advanced computer architectures.
Developed by the IBM Watson Research Center in partnership with many, BlueGene/L was scaled up with a few unique components and IBM's system-on-a-chip technology developed for the embedded microprocessor marketplace. The computer's nodes were interconnected in three different ways instead of the usual one. Using a cell-based design, BlueGene/L was a scalable architecture in which the computational power of the machine could be expanded by adding more building blocks, without introduction of bottlenecks as the machine scaled up.
Key stockpile stewardship application results on BlueGene/L pointed to a qualitative change in the way computational science can be performed. With a rapid time-to-solution, scientists performed a new run every day, made numerous investigations, and explored multiple alternatives. BlueGene/L made it possible for an entire scientific study to be performed in the same time as just one science run required only a year earlier. BlueGene/L made an impact on key NNSA missions and exploring one route toward cost-effective petaflops computing capabilities. It was optimized to run molecular dynamics applications at extreme speeds to address materials aging issues confronting the Stockpile Stewardship Program. BlueGene/L was also a computational science research machine for evaluating advanced computer architectures.
