Gordon Ready to Shatter 35M IOPS
Gordon Ready to Shatter 35M IOPS, Dan Gatti, Big Data I/O Forum
Research is being done at the San Diego Supercomputer Center (SDSC) at the University of California, San Diego investigating solutions in the Big Data explosion. Gordon will be integrated with the new Intel® Xeon® processor E5 Familyand the new Intel® Solid-State Drive 710 Series that will significantly accelerate data intensive applications to achieve up to 35 million(M) Input/Output Operations Per Second (IOPS).
As processor companies increase cores and threads, the Big Data I/O bottlenecks become a larger challenge managing throughput and utilization. The most significant impact will be energy consumption and how to keep utility bills consistent with budgets for all applications processing Big Data.
Gordon employs a vast amount of flash memory to help speed compute solutions now constrained by slower spinning disk technology. This supercomputer deployment is one of the world’s first high performance computing systems to employ massive amounts of Intel® SSD flash memory to greatly reduce I/O latency. Also, new “Supernodes” will exploit virtual shared-memory software to create large shared-memory systems that increase compute speed and yield results faster for the most demanding applications. SDSC has taken delivery of Gordon’s 64 I/O nodes equipped with Intel’s 710 Series, and they are already available to users of Dash, a smaller, prototype version of Gordon.
Gordon is designed for data-intensive applications spanning domains such as genomics, graph problems, geophysics, and data mining. Scientific researches will include the analysis of individual genomes to tailor drugs to specific patients, the development of more accurate models to predict the impact of earthquakes on buildings and other structures, and simulations that offer greater insight into what’s happening to the planet’s climate.
Gordon will be composed of a 1,024 node supercomputer cluster architecture based on the future Intel® Xeon® processor E5 Family. The Intel® processor uses Intel® Advanced Vector Extensions (AVX) to achieve 8 floating point operations/clock cycles providing twice the performance on numerically intensive applications than current processors of the same core count and frequency. The system will be capable of performing in excess of 200 TFlops with a total of 64 terabytes (TB) of memory and 300TB of high performance Intel SSDs served via 64 I/O nodes connected by a dual rail, QDR InfiniBand, 3D torus network. Each of these I/O nodes is capable of more than 560K IOPS, or 35M IOPS for the full system; large memory super nodes that provide up to 2 TB of cache-coherent memory via a high performance software aggregation layer; and access to a 4 petabytes (PB ) Lustre-based file system capable of sustained rates of 100GB/s. Gordon will be factory integrated by Appro’s turn-key System Delivery Integration Services including on-site professional services installation, training and 24/7 maintenance support.
Category: Uncategorized
