HPC news
HIGH-PERFORMANCE COMPUTING
US infrastructure for cyber science sees XSEDE succeed TeraGrid
A partnership of 17 US institutions, led by the University of Illinois’s National Center for Supercomputing Applications, has launched a replacement for the ten- year-old TeraGrid project: the Extreme Science and Engineering Discovery Environment (XSEDE).
Over the past decade, more than 10,000 scientists have used the TeraGrid to complete thousands of research projects in fi elds such as earthquake engineering, materials science, medicine, epidemiology, genomics, astronomy, and biology. ‘Enabling scientifi c discovery through enhanced researcher productivity is our goal, and XSEDE’s ultimate reason for being,’ explained Barry Schneider,
a programme director in the Offi ce of Cyber-infrastructure at the US National Science Foundation. NSF will fund the XSEDE project
for fi ve years, at $121 million. To be truly effective, cyberscience ‘requires a cyberinfrastructure of local computing hardware at sites around the country, advanced supercomputers at larger centres, generally available software packages, and fast networks. Ideally, they should all work together so the researcher can move from local to national resources transparently and easily,’ Schneider added. Initially, XSEDE will support 16 supercomputers across the US.
Europe and Russia to improve oil and gas software
Europe and Russia are to work together towards exascale supercomputing for science and industry. A two-year project, called APOS, will address the programmability challenges of HPC applications in strategically important areas including energy prospecting,
seismic modelling, CFD and nanotechnology. Faster computers do not immediately lead to better results. Incompatibility between existing software and the capabilities of new supercomputers is a growing problem that the new Russian- European collaboration will address.
DATA SPEEDS UP AT US DOE
The US Lawrence Berkeley National Laboratory has taken a major step toward creating one of the world’s fastest scientifi c networks – the Advanced Networking Initiative (ANI) – at a cost of $62 million. Berkeley Lab has signed a subcontracting agreement with Internet2 such that the US Department of Energy’s (DOE) Energy Sciences Network (ESnet) staff will work with Internet2 to
www.scientific-computing.com
develop a 100Gbps prototype network, increasing the information- carrying capacity of DOE’s present scientifi c network by several orders of magnitude. The 100Gbps prototype network is a key step to the DOE’s vision of an eventual 1 terabit network that will support thousands of scientists in research ranging from environmental modelling, developing energy solutions, to exploring the fundamental nature of the universe.
Supercomputer signals new approach to lung cancer
GNS Healthcare is collaborating with the National Cancer Institute (NCI) to accelerate lung cancer research, with a supercomputing platform that can rapidly uncover cause-and-effect mechanisms hidden in huge data sets assembled from imaging, genetics, pathology and other areas. The results could help predict which patients will respond to a given treatment. GNS will analyse NCI data from the
laboratory of Terry van Dyke, director of the Center for Advanced Preclinical Research (CAPR) at NCI. This data was generated from genetically-modifi ed mouse models of non-small-cell lung cancer (NSCLC). This collaboration will utilise GNS’s supercomputer-driven REFSTM platform to build computer models of NSCLC in a hypothesis-free, unbiased manner that will be simulated
to identify key molecular mechanisms of NSCLC. The goal is to identify biomarkers and biological mechanisms that will lead to better matching of drugs to patients and new effective drugs in NSCLC.
Using data from the experimental
assessment of transcriptomic and MRI data relating to NSCLC induction, regression and combination drug treatments, GNS will utilise the REFSTM platform to reverse-engineer network models that connect drug doses to transcriptional and imaging measurement networks, and to endpoints. The results from millions of in silico simulations of these models will provide insights into the fundamental mechanisms of NSCLC and its response to drug treatments, leading to more effective treatments for NSCLC.
GE Global Research acquires supercomputer
GE Global Research has invested in a new Cray XE6m supercomputer. The system will be used to support simulation-based engineering and science across the various disciplines at GE Global Research, and enable the company to run more complex simulations in order to explore multi-physics challenges, gain higher fi delity insights and pursue areas of science and product development that could not be simulated using standard commodity clusters. The Cray XE6m system supports
scalable application workloads and offers a performance-optimised Cray Linux environment. The Cray Cluster Compatibility Mode (CCM) also enables users to benefi t from the streamlined installation of a broad set of independent software vendor (ISV) applications. Using the same petascale technologies as the XE6, the XE6m features the company’s Gemini interconnect and AMD Opteron processors, and is fully upgradeable from previous generation Cray XT5m and Cray XT6m systems.
WEB EXCLUSIVE NEWS now online
A meeting of minds at ISC’11 ISC 2011: The future of GPU accelerators Crunching the numbers
www.hpcprojects.com AUGUST/SEPTEMBER 2011 33
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