Lawrence Livermore Advances in Graph500 Rankings

Print Friendly, PDF & Email
Lawrence Livermore announced at SC14 that its Sequoia supercomputer, a 20-petaflop IBM Blue Gene/Q system, achieved the world’s best performance on the Graph500 data analytics benchmark The Graph500 offers performance metrics for data intensive computing. LLNL and IBM computer scientists attained the No. 1 ranking by completing the largest problem scale ever attempted  (scale 41) with a performance of 23.751 teraTEPS (trillions of traversed edges per second). The team employed a technique developed by IBM.
To fulfill our missions in national security and basic science, we explore different ways to solve large, complex problems, most of which include the need to advance data analytics,” said Dona Crawford, associate director for Computation at Lawrence Livermore. “These Graph500 achievements are a product of that work performed in collaboration with our industry partners. Furthermore, these innovations are likely to benefit the larger scientific computing community.”
In addition to achieving the top Graph500 ranking, Lawrence Livermore computer scientists also demonstrated scalable Graph500 performance on small clusters and even a single node. Livermore computational researchers combined innovative research in graph algorithms and data-intensive runtime systems to achieve these results.
This demonstrates, at two different scales, the ability to solve very large graph problems on modest sized computing platforms by integrating flash storage into the memory hierarchy of these systems. Enabling technologies were provided through collaborations with Cray, Intel, Saratoga Speed and Mellanox.
Our approach really lowers the barrier of entry for people trying to solve very large graph problems,” said Roger Pearce, a researcher in LLNL’s Center for Applied Scientific Computing (CASC).
These results collectively demonstrate LLNL’s pre-eminence as a full service data intensive HPC shop, from single server to data intensive cluster to world-class supercomputer,” said Maya Gokhale, LLNL principal investigator for data-centric computing architectures.
Sign up for our insideHPC Newsletter.