HPC DIRECTOR
forward. I am confi dent that we will reach exascale, and my confi dence is based on experience. In 1995 when DOE conceived the Accelerated Strategic Computing Initiative (ASCI), few computing experts believed the programme would be able to take HPC from approximately 50 gigafl op/s to 100 terafl op/s in under 10 years. The 100-terafl op/s IBM Purple system, dedicated at LLNL in 2005, represented the culmination of that initiative, though it was deployed at the same time as IBM’s fi rst BlueGene/L system, a 360-terafl op/s machine, also at LLNL. BlueGene/L held the No 1 ranking on the Top500 list seven times from 2004 to 2008. ASCI, which brought together the computing expertise of Livermore, Los Alamos, and Sandia national laboratories, established the framework for advancing computing to where it is today with each of the national labs working in partnership with industry to pursue different approaches to realising next- generation supercomputers. We have a lot of experience working with industry to bring new HPC technologies to fruition. In partnership with IBM, we have brought fi ve generations of supercomputers to Livermore and are working to bring in
‘Surmounting these technical and scientifi c obstacles is what is expected of national labs like Livermore’
a sixth next year – the 20-petafl op/s IBM BlueGene/Q system, Sequoia. Previous IBM systems include: the 3-terafl op/s Blue Pacifi c; the 12-terafl op/s White; the 100-terafl op/s Purple; the 500-terafl op/s BlueGene/L; and the recently installed 500-terafl op/s BlueGene/P system called Dawn, which is paving the way for Sequoia. With Sequoia we are making the single biggest leap ever attempted in computing power, from 500 terafl op/s today to 20 petafl op/s in 2012. To reduce the time required to adapt applications to multi-petafl op/s systems and to develop the network and storage these systems will require, we have again turned to industry for assistance.
Through an innovative collaborative project called Hyperion, we are working with 10 computing industry partners to address some of these issues. Hyperion brings together
Dell, Intel, Supermicro, QLogic, Cisco, Mellanox, DDN, Sun, LSI, and RedHat to create a large-scale testbed (100 terafl op/s) for HPC technologies critical to our work maintaining the aging US nuclear weapons stockpile, as well as industry’s ability to make petafl op/s computing and storage more accessible to commerce, industry, and research and development.
Hyperion represents a new way of doing
business. Collectively, we have been able to build a testbed none of us could have built individually. This project is a cost-effective way to advance the state-of-the-art while benefi tting both the end users, such as national labs like Livermore, and industry. Building on these collaborations and forming innovative partnerships will be essential to achieving exascale computing. The challenge is daunting and the distance appears beyond bridging. But to those of us at Livermore who have been part of this computational enterprise for a while, this position is not unfamiliar.
Surmounting these technical and scientifi c obstacles is what is expected of national labs like Livermore. And we eagerly pick up the gauntlet.
Create an Analytical Knowledge Management System
•Vendor neutral and multi-technique support for NMR, MS, Chromatography, UV, IR, and more
•Fast and versatile searching by structure, spectrum, chromatogram, or text-based queries
•Simple and flexible reporting supports collaboration and decision-making • Incorporate data into existing LIMS and ELNs • Add automation capabilities to help improve efficiency
• Scalable to suit small research groups, mid-sized labs, or global organizations
Learn more about ACD/SpecManager
www.acdlabs.com/specman/
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33 |
Page 34 |
Page 35 |
Page 36 |
Page 37 |
Page 38 |
Page 39 |
Page 40 |
Page 41 |
Page 42 |
Page 43 |
Page 44 |
Page 45 |
Page 46 |
Page 47 |
Page 48