Labs share credit for award-winning IBM Blue Gene series

Posted In: R&D Daily | Modeling | Simulation

Thursday, October 8, 2009

Loading...



 

The IBM Blue Gene series of energy-efficient supercomputers, central to breakthrough scientific research around the world, were singled out by President Barack Obama as a Medal of Technology and Innovation award-winner on October 7 in Washington, D.C.

When IBM accepts this well-deserved honor, computer scientists at the U.S. Department of Energy’s (DOE) Argonne and Lawrence Livermore national laboratories can take pride in their role in making these computers a reality. Both sites contributed critical input and software components through a DOE research and development partnership with IBM that strongly impacted Blue Gene’s extreme-scale design.

“The success of this partnership is an excellent example of how national laboratories can help fuel industry and drive innovation ,” said William Brinkman, Director of DOE’s Office of Science.

“The Blue Gene supercomputers are an outstanding example of our investment in nuclear security providing the tools to tackle broader national challenges,” said NNSA Administrator Thomas D’Agostino. “This machine, which was originally developed to ensure the safety and reliability of our nuclear stockpile without testing, has led to amazing advances in science and discovery. I congratulate IBM, Lawrence Livermore National Laboratory and Argonne National Laboratory for a job well done.”

DOE, which leads the world in providing supercomputers for scientific research, began an R&D partnership with IBM in 2001 to develop the Blue Gene platform. Argonne, together with Lawrence Livermore National Laboratory (LLNL), worked on key aspects of the design and provided critical software components to ensure it was well suited to solve challenges in energy, the environment and national security.

“Blue Gene balances energy efficiencies with a major breakthrough in scalability, which is imperative to attack problems in science and engineering at unprecedented scale and speed,” said Pete Beckman, director of Argonne’s Leadership Computing Facility, home to the first large Blue Gene/P.

The Blue Gene leverages a high-performance, low-power, system-on-a-chip architecture offering extreme integration that dramatically improves reliability increases energy efficiency and reduces operating costs.

“The Blue Gene architecture has greatly enhanced the National Nuclear Security Administration’s capability for predictive simulation and uncertainty quantification,” said Mark Seager, Asst. Dept. Head for Advanced Computing Technology at LLNL.

Much of the software needed to operate Blue Gene comes from the open source community and was developed by laboratories and universities around the world. Argonne was actively involved in fostering that community as well as developing key components of the system software. For example, the Blue Gene leverages Argonne's MPICH, the version of the Message Passing Interface that scientists use to write parallel programs capable of scaling to hundreds of thousands of CPU cores. Computer scientists are also working on extending the capabilities of Blue Gene with advanced math libraries, improved parallel file systems, and even experimental operating systems such as ZeptoOS, which permits users to run Linux on the Blue Gene’s compute nodes.

Another critical aspect of the Blue Gene’s success was ensuring the platform was adopted by the high-performance computing community. In 2004, Argonne and IBM jointly created the Blue Gene Consortium, an international group of laboratory, university and industrial researchers collaborating to evaluate the technology and platform and provide critical feedback for future Blue Gene designs.

The Medal of Technology and Innovation is described as the highest honor for technological achievement for outstanding contributions related to the nation’s competiveness, standard of living and quality of life through the development and commercialization of technological products. Obama will present the medal to IBM, as well as four inventors who were also recognized for the award, at a White House ceremony on October 7. The medal program is administered by the United States Patent and Trademark Office and is awarded annually to individuals, teams and companies.  

“It is an honor and a privilege to be part of this partnership,” said Beckman. “The Blue Gene/P is already making a measurable impact on the research community and enabling advancements that will shape our future”.

IBM Research Blue Gene page

SOURCE: Argonne

JOIN THE DISCUSSION
Rate Article:  Average 0 out of 5
Register or log in to comment on this article!

0 Comments

Add Comment

Text Only 2000 character limit

Page 1 of 1

New To Market

more

P2i showcases liquid repellent nano-coating for hearing aids
P2i showcases liquid repellent nano-coating for hearing aids

At the AudiologyNOW! 2010 show in San Diego next month, UK-based coatings company P2i will display their relatively new Aridion liquid-repellant nano-coating. Designed for exposure to humidity or sweat, the polymer layer is applied by a pulsed ion gas process that lower’s the hearing aid’s surface energy, coaxing water away from delicate components.

Submersible FlowCAM catches particle images and data in-situ and real-time

Fluid Imaging Technologies recently introduced its Submersible FlowCAM particle and cell imaging and analysis system at Ocean Sciences 2010 in Portland, Ore. The remote sensing platform can be used for continuous, unattended monitoring tethered to research vessels or autonomous submersibles.

Tools & Technology

more

Benchtop NMR analyzer
Benchtop NMR analyzer

Oxford Instruments America, Inc.’s Magnetic Resonance Group released the second generation of its MQC analyzers.

Software solution for microarray image analysis

BioDiscovery Inc. released ImaGene 9.0 for microarray image analysis. The new features include improved memory performance for the latest high density arrays, streamlined processing pipeline focused on image quantification and intensity extraction, and new modular design with options to add modules for analysis of gene/miRNA expression or CGH data.

Advertisement

Advertisement