Ranger Supercomputer Dedicated by NSF and Texas Computing Center

Tuesday, February 26, 2008

Quantum leap in computational power and memory capacity for research community through the TeraGrid

Understanding HIV drug-resistance (click image for larger version) More Information ROLLOVER Credit: Peter Coveney, University College London. Texas Advanced Computing Center.

Modeling Mantle Convection (click image for larger version) More Information ROLLOVER

"Ranger is the first leadership computational resource provided under the National Science Foundation ‘Track 2 initiative' and the first machine funded through the newly formed Office of Cyberinfrastructure," he said. The Track 2 initiative is NSF's four-year activity designed to fund the deployment and operation of up to four leading-edge computing systems that will greatly increase the availability of computing resources to U.S. researchers. The Ranger award, the largest NSF grant to the University of Texas at Austin, was made after the evaluation of selection criteria that were "multi-faceted, including not only raw performance of the machine, but also effective education and outreach commitments, institutional competency and commitment for service to the national research community," noted Atkins.

"Ranger is a dramatic step forward in offering computational resources for the open science research community," he said, stressing that computational and data processing resources such as those provided by Ranger are now absolutely essential to frontier research in almost every field of science and engineering. He called Ranger and associated visualization services "a tool for finding knowledge 'needles' in enormous digital data 'haystacks,' and then integrating these needles into spires of insight and breakthrough discovery."

Ranger offers more than six times the performance of the previous largest system for open science research. The boost in performance offered by Ranger relative to the previously largest open science machine is comparable to reducing the flight time from New York to London to just one hour.

Atkins provided context for the new Ranger facility, as a part of a larger network of integrated resources available for the nation's research communities known collectively as the "TeraGrid." "Ranger brings a quantum leap in computational power and memory capacity to the TeraGrid and is a major step forward towards the next goal of computing at sustained rates of 10 to the 15^th operations per second-at the petascale level," he said

Ranger's deployment marks the beginning of the Petascale Era in high-performance computing (HPC) where systems will approach a thousand trillion floating point operations per second and manage a thousand trillion bytes of data.

Ranger is the largest HPC computing resource on the NSF TeraGrid, a nationwide network of academic HPC centers that provides scientists and researchers access to large-scale computing power and resources. Ranger will provide more than 500 million processor hours of computing time to the science community, performing more than 200,000 years of computational work over its four-year lifetime.

Ranger and other petascale systems to follow will address many of society's most pervasive grand challenges including global climate change, water resource management, new energy sources, natural disasters, new materials and manufacturing processes, tissue and organ engineering, patient-specific medical therapies, and drug design. These issues cannot be addressed or overcome without modeling and simulation.

Ranger is a collaboration among TACC, The University of Texas at Austin's Institute for Computational and Engineering Sciences (ICES), Sun Microsystems, Advanced Micro Devices, Arizona State University and Cornell University. The $59 million award covers the system and four years of operating costs.

The full text of Atkins' remarks follows:

Good afternoon. This is a very exciting day for 21st century research and education. This day has been made possible by the cooperative efforts of many people and numerous organizations. I will be more specific about this shortly, but first some context.

Please let me begin at the personal level. This is an emotional day for me because it makes vivid the compounding, exponential advances in computing technology in less than 50 years and the enormous privilege I have had to participate in the incredible growth of performance, impact, and ubiquity of digital computing technology from the days of its infancy. As a graduate student at the University of Illinois I was part of the one of the first transistorized computer projects - Illiac 2. It was built from discrete transistors costing $80 a piece in (1960 dollars), occupied a room larger than the Ranger facility, and provided much less compute power than one of the AMD CPU cores used in Ranger. And in those days the word "core" referred to a ferrite donut providing one-bit of storage in the computer's main memory. I could have never imagined that it would become the label for a tiny complete central processor building block in today's massively parallel multi-cores machines.

Ranger is the first leadership computational resource provided under the National Science Foundation "Track 2 initiative" and the first machine funded through the newly formed Office of Cyberinfrastructure. This award was made under a revised policy at NSF that emphasizes a holistic notion of "impact on science" as the overarching criteria for making cyberinfrastructure awards. This science-driven award process involves close cooperation between the experts in providing high performance computation resources, and the science and engineering research communities who must now have these resources to pursue transformative discovery and learning. Benchmarks as surrogates of the algorithm-types important to research communities are developed and the awardees are expected to meet or exceed them. The criteria are multi-faceted including not only raw performance of the machine, but also effective education and outreach commitments, institutional competency and commitment for service to the national research community.

These Track 2 awards are also made under what NSF calls a "cooperative agreement." The awardee and NSF cooperate in developing a plan of action, consistent with the submitted proposal, with mutually agreed performance criteria and milestones along the way. They then cooperate in realizing the plan. In the case of Ranger, I was very pleased to learn that this spirit of cooperation has extended generously to the relationship between TACC and Sun Microsystems and AMD. I have been told that you have all conceived this as a collective opportunity and gone the extra mile to insure that Ranger is a dramatic step forward in offering computational resources for the open science research community. The dedication of Ranger today and all that has lead up to it has set a very high standard in both process and performance for the entire future of HPC provisioning by NSF. You have set a very high bar.

Computational and data processing resources such as provided by Ranger are now absolutely essential to frontier research in almost every field of science and engineering. Computationally-based modeling, simulation, and prediction across huge scales of time and space are the norm in exploring emergent complex systems such as embedded in social and physical global climate change models, understanding and predicting turbulent weather, understanding of the universe, the fundamentals of life, and global patterns of social interaction. Computational science is now a third mode of discovery complementing theory and physical experimentation. Science and engineering research is also increasingly data-intensive due in large part to enormous breakthroughs in sensing and observing technologies. Ranger and associated visualization services are also a tool for finding "knowledge needles" in enormous "digital data haystacks" and then integrating these needles into spires of insight and breakthrough discovery.

So this is indeed a grand day for celebration and appreciation. And now back to recognizing the team that has brought us to this day.

I begin by thanking

• University of Texas President William Powers for his and his administration's enlightened support of this world-class leadership facility and the culture of contribution to our nation it exhibits.

• I thank Jonathan Swartz and Sun Microsystems for their truly innovative machine and their willingness to go the extra mile in service to our nation's research communities.

• I thank Hector Ruiz and AMD for their leadership class, multi-core processors (used here and in other NSF-supported facilities) and their also "extra-mile" mind set and strong commitment to making Ranger a superlative resource.I thank Congressman McCaul and the U.S. Congress for the appropriations to NSF that provides us the opportunity to invest in our Nation's future capacity to compete in our increasingly competitive, knowledge-based world.

• I thank Jay Boisseau and his team at TACC for their extraordinary competency and dedication to provide unique computational resources to the Nation.

• And I also thank the TACC academic partners

  • The Institute for Computational Engineering and Sciences at UT directedby Tinsley Oden;

  • the Arizona State High Performance Computing Institute, directed by Dan Stanzione,

  • and the Cornell Center for Advanced Computing directed by Dave Lifka.

In the spirit of Ranger being a team effort, I also acknowledge and thank Steve Meacham and Jose Munoz, members of the NSF OCI team, for structuring and presiding over a very high quality peer review process and an award recommendation that led to the support for Ranger, and for their ongoing dedication to making this investment a huge success.

Although we focus today on this new Ranger facility, it is important to realize that it is part of a larger network of resources known collectively as the TeraGrid. The TeraGrid is a consortium of network resources increasingly integrated as a collective resource for the nation's research communities. Ranger brings a quantum leap in computational power and memory capacity to the TeraGrid and is a major step forward towards the next goal of computing at sustained rates of 10 to the 15th operations per seconds - i.e. at the petascale level. The TeraGrid enables researchers to link and use multiple computers with diverse architectures, data bases, and visualization facilities independent of where they are and who operates them. Increasingly TeraGrid resources are being used transparently through Web-based science gateways that are tailored to both research and education in a growing number of fields. A high school student can design and analyze nano-structures using the same tools as the researchers, or even make original contributions to astronomy through data mining in a digital sky. These science gateways are entrances into virtual communities that enable participation in authentic discovery and learning independent of time and place. They open new opportunities in science for what we do, how we do it, and who participates.

Thank you all for your past and future contributions to enabling discovery and innovation essential to the well-being and vitality of our world. Congratulations and all the best. If now is the time for applause, let me ask that it be hearty applause for the entire team that worked so hard to bring us to this day of celebration.

Source: NSF