For more information please send us your email address.
TECHFIND - DATAPROCESSING
Gordon Supercomputer
A supercomputer, which uses flash memory instead of traditional hard drives, is being developed at a California University under a grant from the National Science Foundation. The flash memory will be Intel High Performance solid-state drives. The system will allow researchers to tackle a growing list of critical data-intensive problems such as analyzing individual genomes to tailor drugs to specific patients; developing more accurate models to predict the impact of earthquakes on buildings and other structures; and in climate simulations. The system will use 32 'supernodes' that will exploit virtual shared-memory software to create large shared-memory systems in order to make computation faster. Each supernode will consist of 32 computing nodes, capable of 240 gigaflops per node - one gigaflop or GF equals a billion calculations per second - and 64 gigabytes of digital random access memory. The supernode will also have two I/O nodes, each with 4 terabytes of flash memory. Linked through virtual shared memory, each of the 32 supernodes will have the potential of 7.7 teraflops of computing power and 10 TB of memory.
Media Spiders®
A supercomputer, which uses flash memory instead of traditional hard drives, is being developed at a California University under a grant from the National Science Foundation. The flash memory will be Intel High Performance solid-state drives. The system will allow researchers to tackle a growing list of critical data-intensive problems such as analyzing individual genomes to tailor drugs to specific patients; developing more accurate models to predict the impact of earthquakes on buildings and other structures; and in climate simulations. The system will use 32 'supernodes' that will exploit virtual shared-memory software to create large shared-memory systems in order to make computation faster. Each supernode will consist of 32 computing nodes, capable of 240 gigaflops per node - one gigaflop or GF equals a billion calculations per second - and 64 gigabytes of digital random access memory. The supernode will also have two I/O nodes, each with 4 terabytes of flash memory. Linked through virtual shared memory, each of the 32 supernodes will have the potential of 7.7 teraflops of computing power and 10 TB of memory.
