Where to turn in a time of need. In need of MFLOPS, that is.
"The only constant is change" said Heraclitus, an ancient Greek philosopher. This is certainly true when describing the market of high-performance computing (HPC). The HPC market is characterized by a rapid change of architectures, technologies and usage. The only continuously steady characteristic of HPC is the ever-growing demand for performance, showing an increase of 100X every ten years based on the TOP500 list. The HPC market drives the computing technology further and evaluates many leading edge architectures. Proven solutions that have been widely adopted in the HPC are accepted in commercial high-performance computing markets such as automotive, oil and gas, financial, digital media, bio-science and mathematical modeling.
Clustered commodity servers have become the dominant solution for HPC systems as they offer tremendous price/performance benefits, unparalleled flexibility in deployment, and reduced long-term maintenance. With the fast adoption of the cluster architecture, the importance of integrating all of its parts is paramount, from hardware â CPU, memory, interconnect and attached storage â to software, including drivers, libraries (such as MPI), operating systems and parallel applications.
While high-performance computing continues to be one of the fastest-growing segments in the IT industry, the general commercial sector is still underutilizing the available HPC technologies. In many cases, users do not maximize the potential of high-performance computing to get better, safer, and more reliable products to market more quickly. Moreover, in some cases, users have not made the jump from single workstations into HPC clustering due to lack of knowledge, even though they are facing problems that can not be solved on their current systems.
The HPC Advisory Council is a computing ecosystem that includes best-in-class original equipment manufacturers (OEMs), strategic technology suppliers, independent software vendors (ISVs) and selected end-users across the entire range of HPC market segments. As of September 2008, the Council includes more than 50 members, including: AMD, Appro, Blue Ridge Numerics, Clustercorp, Colfax International, Corning Cable Systems, DataDirect Networks, Dell, Evergrid, Fermi National Accelerator Laboratory, GigaSpaces Technologies, HCL Infosystems, HP, IBRIX, IBSwitches.com, Intel, The Israeli Association of Grid Technologies (IGT), Lamprey Networks, Livermore Software Technology Corporation, LSI Corporation, Mellanox Technologies, Microsoft, Microway, NEC Corporation of America, Netweb Technologies, Network Equipment Technologies, Oak Ridge National Laboratory, Ohio State University, Panasas, ParTec Cluster Competence Center GmbH, RNA Networks, SGI, Scalable Graphics, ScaleMP, Schlumberger, Silicon Mechanics, SoftModule, Sun Microsystems, Swiss National Supercomputing Centre CSCS, System Fabrics Works, Terascala, The Victorian Partnership for Advanced Computing, Virginia Tech University, Voltaire, VXTECH, W.L. Gore & Associates, Wipro InfoTech, Wolfram Research and Z Research.
Along with the HPC Cluster Center, the HPC Advisory Council targets education and outreach as one of its main goals. Together, the Council educates users on how to migrate from workstations or desktops into high-performance clustering in order to achieve productivity and capabilities that deliver better and more innovative products while also bringing them to market more quickly. The Council will provide recipes for building efficient clustering solutions, will post technical information, and will drive real case studies to be performed by end-users with the Councilâs support.