Panasas ActiveScale Storage Cluster Will Provide I/O for World’s Fastest Computer: Panasas today announced that Los Alamos National Laboratory (LANL) will be deploying the Panasas ActiveScale Storage Cluster as the storage solution of choice for its Roadrunner petascale supercomputer, which is intended to become the fastest computer in the world. LANL will expand its current installation of the Panasas Storage Cluster to provide a leading edge storage solution for the new petaflop system. Roadrunner will run complex, scientific calculations based on the Linux operating system and the Panasas Storage Cluster with DirectFLOW. The DirectFLOW capability offers users such as LANL with a fully-parallel data path to allow high speed, direct communications between their Linux cluster and Panasas storage. The Roadrunner petascale supercomputer is a radical departure from the existing top 10 supercomputers in that it will be the largest supercomputer ever built based on a hybrid architecture. The machine will utilize AMD Opteron general purpose processors and the new IBM Cell Broadband Engine processor, originally designed for video game platforms. The supercomputer will be used to run nuclear weapons calculations for the Department of Energy’s National Nuclear Security Administration. This endeavor will help to usher in an era of a new computing paradigm where hybrid architectures are used for extreme scale computation. Currently, LANL is using the Panasas Storage Cluster to provide a common globally-shared, parallel file system to seven terascale supercomputers. The Panasas Storage Cluster at LANL was the first parallel file system to be shared globally among multiple terascale clusters at data rates in the tens of gigabytes-per-second range. LANL has more than one petabyte of Panasas storage that provides a total of approximately 50 gigabytes per second of bandwidth. The existing Panasas infrastructure will grow substantially over the next few years to accommodate the new Roadrunner petascale system. “We’ve been using Panasas storage for a long time at LANL to provide scalable and globally-shared storage to multiple terascale clusters. We will leverage our successful, scalable, and stable Panasas storage solution to provide the I/O solution for the Roadrunner system,” said Gary Grider, group leader of Los Alamos’ High Performance Computing Systems Integration Group. “The Roadrunner system with its hybrid architecture will be a big challenge for LANL and the entire supercomputing community. It makes sense for us to leverage our existing and highly productive Panasas solution for LANL’s Roadrunner system.” The Panasas Storage Cluster is an integrated hardware and software solution that combines an object-based architecture with a parallel file system. The intelligent Panasas storage technology is easy to install and manage including substantial self-management capabilities that allow IT administrators to be more productive. By delivering scalable capacity and high performance from within a single global namespace, the Panasas Storage Cluster enables LANL to optimize cluster application performance and maximize investments already made in the lab's HPC environment. "Panasas has a long history of working with Los Alamos in implementing cutting-edge computing technologies. As the transition begins in earnest to the next frontier in high performance computing, the Panasas Storage Cluster will continue to deliver the performance, scalability, reliability, and manageability that these petascale supercomputers will need," said Larry Jones, vice president of Marketing at Panasas.