HPCwire: NY Supercomputer 'Magic' Available Worldwide

King Abdullah University of Science and Technology

Since 1986 - Covering the Fastest Computers
in the World and the People Who Run Them

Translation Disclaimer

Subscribe | Sign In

Visit additional Tabor Communication Publications

Leading HPC
Solution Providers

April 06, 2009

NY Supercomputer 'Magic' Available Worldwide

HWI, UB scientist in partnership with NVIDIA and Dell rolls out "Magic" to worldwide users

BUFFALO, NY, April 6 -- Local scientist Dr. Russ Miller is leading the rollout of "Magic," one of the most powerful computers in New York State to qualified users worldwide for solving computationally-demanding problems.

New Methods Allow Scholars to Address Previously Unsolvable Problems

Cyberinfrastructure sits at the core of modern simulation and modeling, which creates new methods of investigation that allow scholars to address previously unsolvable problems, according to Miller who holds scientific appointments at both Hauptman-Woodward as a senior scientist and at the University at Buffalo as a Distinguished Professor of Computer Science and Engineering. The Miller Cyberinfrastructure Laboratory (MCIL) is taking advantage of recent advances in technologies in order to link distributed resources, including compute systems, data storage devices, visualization systems, sensors, and a wide variety of instruments and make those resources available worldwide.

"The work done by Dr. Miller's group, which was supported by an NSF grant that also providing funding to Niagara University (NU) provided us with the opportunity to acquire significant computing power and skills. It also allowed us to train students on creating and utilizing modern computing platforms," Dr. Mary McCourt, chair and professor of Chemistry at NU, said. "We are excited to have immediate access to this new state-of-the-art computing system, which enhances the resources of Niagara's Academic Center for Integrated Sciences."

"We have been working on grid computing efforts with Russ and his group for years," Dr. Charles M. Weeks, senior research scientist at HWI, said. "These efforts have resulted in a number of joint research and funding efforts. We are particularly anxious to experiment with this new trend in high-performance computing as part of one of our jointly funded projects, especially for some of HWI's computationally demanding applications."

Project History

The Miller Cyberinfrastructure Laboratory (MCIL) was founded at the beginning of the 21st century by Miller who is widely recognized as a leader in parallel computing, grid computing, and cyberinfrastructure. Miller served as Founding Director of the UB Center for Computational Research (CCR) from 1998-2006. Under Miller's direction, and with primary funding from the National Science Foundation (NSF) and National Institutes of Health (NIH), a Buffalo-based grid was developed as an experimental way to create an institutionally-distributed platform that integrated information and computing. This led to the design, implementation, and deployment of the Western New York Grid (WNY Grid), which aggregated costly computational resources at institutions throughout WNY. The overwhelming success of the WNY Grid led to the MCIL-deployed New York State Grid (NYS Grid), based on the Open Science Grid set of grid software.

Magic is a state-of-the-art computer system located at UB's North Campus which consists of graphics processing units typically reserved for high-end gaming systems that are integrated into a traditional rack of computers. Magic was delivered in late 2008 and is now available to users worldwide. The system is cost-effective in solving large computational problems in areas including bioinformatics, computational chemistry, computational fluid dynamics, computational finance, medical imaging, weather and ocean modeling.

Why Is This Important?

Led by Miller, MCIL worked on platforms for monitoring grid systems, for providing a single point-of-entry portal and for solving intricate problems involving node swapping, predictive scheduling and resource management. MCIL has been responsible for grid-enabling critical applications in areas such as structural biology, bioinformatics, ground water modeling, earthquake engineering and computational chemistry.

"The new partnership with NVIDIA and our ongoing partnership with Dell have made possible the acquisition of a leading-edge machine, one of the fastest in New York for solving a large set of scientific, engineering and multimedia problems," Miller said. "In fact, due to the advances in heterogeneous computing, this machine cost approximately 1/1000th of the two machines in New York that have peak computing power exceeding this machine. In particular, Graphics Processing Unit (GPU)-based systems are affordable by many research groups and departments offering processing capabilities thousands of times faster than traditional systems, whereas traditional high-end clusters require extraordinary funds, space, cooling, and staffing. This is also the fastest machine on the MCIL Grid for solving many computationally-intensive problems."

Professor Jack Dongarra, one of the foremost authorities on high-end computing and director of the Innovative Computing Laboratory at the University of Tennessee said, "GPUs have evolved to the point where real-world applications are easily implemented on them and run faster than on multicore systems. Future computing architectures will be hybrid systems with parallel-core GPUs working in tandem with multicore CPUs."

"The Tesla GPU-based clusters used by the Cyberinfrastructure Laboratory are enabling its researchers to get supercomputing performance from smaller clusters at 1/100th the cost, significantly improving the scope and size of the research problems that they can solve. More importantly they can do this while consuming less electricity, reducing the carbon footprint required to power these systems and cool the data centers that house them," Andy Keane, general manager, GPU Computing at NVIDIA, said.

The NVIDIA-based GPU-system was acquired with funds from the National Science Foundation's (NSF's) Collaborative Research Infrastructure (CRI) Program based on a joint proposal from UB (Miller), Niagara (McCourt), SUNY-Geneseo (Farian), and the Hauptman-Woodward Institute (Weeks). These funds have been used previously to acquire compute clusters at each of these institutions to support research, scholarship, and education at and between these institutions. In fact, at Niagara University, the Academic Center for Integrated Sciences, which will be housed in the new Thomas B. Golisano science building, has benefitted immensely from this collaboration in terms of intellectual and capital resources. The system was brought on-line by Kevin Cleary, systems administrator in the UB Department of Computer Science and Engineering, with guidance and assistance from Jon Bednasz, senior systems administrator in the UB Center for Computational Research.

About the Collaborators

For additional information, contact Miller at miller@buffalo.edu. The Hauptman-Woodward Medical Research Institute is an independent, not-for-profit, biomedical research facility located in the heart of downtown Buffalo's medical campus. HWI scientists study the 3-D structures of protein molecules that play a role in the development of many different diseases. Niagara University was founded in 1856 and has been in existence for 150 years and enrolls more than 3800 students. The University at Buffalo is a premier research-intensive public university. It is the largest and most comprehensive campus in the SUNY system. Information on UB's Department of Computer Science and Engineering is available at www.cse.buffalo.edu. Information on CCR is available at www.ccr.buffalo.edu. NVIDIA is the world leader in visual computing technologies and the inventor of the GPU, a high-performance processor which generates interactive graphics on workstations, personal computers, game consoles, and mobile devices. Open Science Grid information is available at www.opensciencegrid.org.

-----

Source: The Hauptman-Woodward Medical Research Institute

Share Options

Subscribe

Subscribe to HPCwire

Discussion

There are 0 discussion items posted.

Join the Discussion

Become a Registered User Today!

Registered Users Log in join the Discussion

RSS Feeds

Feeds by Topic

Feeds by Industry

Feeds by Content Type

Subscribe to All Content

All

May 23, 2012

May 22, 2012

May 21, 2012

May 18, 2012

May 17, 2012

May 16, 2012

May 15, 2012

May 14, 2012

May 11, 2012

Feature Articles

NVIDIA Works On CPU Co-Dependency Issues with Kepler GPU

NVIDIA is telling everyone that the GK110, its new Kepler GPU aimed at supercomputing, is all about improving performance per watt. But the other driving theme behind the new architecture is reducing the GPU's reliance on its CPU host. How well it accomplishes both these goals areas could determine the success of the new chip in high performance computing.
Read more...

OpenACC Starts to Gather Developer Mindshare

PGI, Cray, and CAPS enterprise are moving quickly to get their new OpenACC-supported compilers into the hands of GPGPU developers. At NVIDIA's GPU Technology Conference this week, there was plenty of discussion around the new HPC accelerator framework, and all three OpenACC compiler makers, as well as NVIDIA, were talking up the technology.
Read more...

NVIDIA Launches Kepler Into HPC

NVIDIA has introduced its first Kepler-generation GPU product for high performance computing, and revealed some of the inner working of the new architecture. The announcement took place at the kickoff of the company's GPU Technology Conference taking place this week in San Jose, California.
Read more...