July 01, 2010
Here is a collection of highlights from this week's news stream as reported by HPCwire.
Massively Parallel Technologies Awarded Patent for Scalable Application Environments
Australia's Victorian Life Sciences Computation Initiative Deploys SGI Cluster
Carnegie Mellon Software Engineering Institute Receives Contract Extension
Microsoft, Intel to Support Local College Internship Program with Avetec's DICE Program
Berkeley Lab Announces New International Center for Computational Science
IBM and Slovenia's Arctur Bring Supercomputing to the Midmarket
Oracle Unveils Next Generation Sun Fire x86 Clustered Systems
Ohio Supercomputer Center, Sciences Computer Consultants Forge Alliance
PBS Professional Roll Now Available to Clustercorp's Rocks+ Community
Verari Gains Momentum with New Investment, Leadership
NERSC, HDF Group Optimize HDF5 Library to Improve I/O Performance
Partnership of Supercomputing Centers Announce NSF XD TIS Award
Digipede Launches Digipede Network 2.4
NASA Goddard Introduces the NASA Center for Climate Simulation
Mellanox InfiniBand Delivers Record Message Rate Performance
Green500 List Gets a Little Greener
The latest Green500 list was announced by its overseers at Virginia Tech, giving due praise to the machines that perform the best whilst consuming the least energy, an energy-efficiency equation the Green500 defines as 'millions of floating-point operations per second' (MFLOPS) divided by 'watts' (W), or MFLOPS/W. This time around, accelerator-based supers have a prominent position on the list, where they occupy the top eight spots. Green500 Co-Founder Wu Feng, who is also an associate professor of Computer Science at Virginia Tech, compared the performance of accelerator-based computers with traditional supercomputers:
"The accelerator-based supercomputers on The Green500 List produce an average efficiency of 554 MFLOPS/W whereas the other measured supercomputers on the list produce an average efficiency of 181 MFLOPS/W. That makes the accelerator-based supercomputers on the Green500 more than three times more energy efficient than their non-accelerated counterparts on the list."
This increase in accelerator-based systems has led to an upswing in the overall efficiency of the list to its current average of 140.33 MFLOPS/W -- a 12.5 percent increase since November (124.71 MFLOPS/W average) and 29 percent increase since last June (108.83 MFLOPS/W average). Here's a rundown on the top 10 finishers (there are multiple ties throughout):
1 Forschungszentrum Juelich, QPACE Cluster
1 Universitaet Regensburg, QPACE Cluster
1 Universitaet Wuppertal, QPACE Cluster
4 National Supercomputing Centre in Shenzhen (NSCS), Nebulae
5 DOE/NNSA/LANL, BladeCenter Cluster
5 IBM Poughkeepsie Benchmarking Center, BladeCenter Cluster
7 DOE/NNSA/LANL, BladeCenter Cluster
8 Institute of Process Engineering, Chinese Academy of Sciences, Mole-8.5 Cluster
9 Mississippi State University, iDataPlex
10 Banking (M), iDataPlex
This is the seventh edition of the Green500 list, which made its debut in November 2007. The list comes out shortly after each TOP500 list, which is announced twice a year, during SC in November and during ISC in June. Performance data used in creating the list are obtained from publicly-available sources including the TOP500 list.
The June 2010 Green500 list is available here.
Goodbye EGEE, Hello EGI
As of April 30, 2010, the Enabling Grids for E-SciencE (EGEE) is no more, but it is leaving behind a considerable legacy and setting the stage for grid collaborations to come. The organization released an announcement today to highlight some of their biggest achievements and to set the stage for the handoff to the next instantiation of the European grid schema: the European Grid Infrastructure, EGI.
EGEE grew out of the European DataGrid project, which commenced in 2001. There were three successive phases of EGEE, culminating in EGEE-III, which brought together the infrastructure, software tools and services to support more than 10,000 researchers across more than 170 research communities. In fact, EGEE can boast the largest collaborative production grid infrastructure in the world, made up of 250,000 computer processing cores, hosted by more than 300 computing centers, and running approximately 15 million computer tasks every month.
EGEE Project Director Bob Jones considers the history of the collaboration and how the project will live on in the body of its successor, EGI:
"Distributed computing infrastructures have matured at an incredible rate and EGEE has been a driving force in making this happen. In 2000 I think it would have been a brave call to say that Europe would have a sustainable production grid infrastructure built on a federation of national grid infrastructures in place by 2010. We are glad to see EGI embody the results of a decade of work and I am certain it has a bright future in contributing to the European Research Area and serving the ESFRI projects."
The new project, the European Grid Infrastructure, EGI, brings together the federation of individual National Grid Infrastructures (NGI) under the coordination of a new organization, EGI.eu. The European Commission is providing initial co-funding for EGI through the four-year EGI-InSPIRE project, which began May 1, 2010, also coordinated by EGI.eu. The EGI-InSPIRE project will support the transition from a project-based system to a sustainable pan-European e-Infrastructure, by supporting "grids" of high-performance computing (HPC) and high-throughput computing (HTC) resources. EGI-InSPIRE will also integrate new distributed computing infrastructures such as clouds, supercomputing networks and desktop grids.
According Steven Newhouse, EGI.eu Director, sustainability will be a key focus:
"The establishment of EGI.eu represents a new phase for the European Grid Infrastructure. Sustainability is key for supporting the next generation of data-intensive science projects. EGI, through its foundation on strong National Grid Infrastructures, will provide the coordination necessary for a secure, reliable and integrated infrastructure in Europe."
As part of its wrap-up process, EGEE has created a final project review, which took place in June. The final project summary can be accessed here (PDF).
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?