Congressman Dan Lipinski (IL-3) joined with U.S. Department of Energy (DOE) Under Secretary for Science and Energy Lynn Orr, Argonne National Laboratory Director Peter Littlewood, and other members of the Illinois congressional delegation to announce that Argonne is receiving $200 million to increase its supercomputing capacity. This investment will help advance U.S. leadership in scientific research, maintain the nation's position at the forefront of next-generation computing, and ensure America's economic and national security. Rep. Lipinski helped write the 2010 America COMPETES Act that increased investments in computing and other innovation programmes.
"Supercomputers like MIRA at Argonne have traditionally been a strength of the U.S. investment in research, but today that lead is being challenged by other countries", stated Rep. Lipinski. "While the U.S. is home to 233 of the world's 500 fastest computers as measured by the most recent TOP500 report, that's down from 291 systems just five years ago. This reward for Argonne is critical to our nation's ongoing efforts to maintain our lead in this sector. I thank the DOE and Argonne Director Littlewood for their leadership and vision in keeping the United States at the forefront of this highly competitive field. As a member of the House Science, Space, and Technology Committee, the Ranking Member of the Subcommittee on Research, and a member of the Subcommittee on Energy, fighting for increased investments for high performance computing and other critical areas of science is going to continue to be at the top of my priority list."
Under the joint Collaboration of Oak Ridge, Argonne, and Lawrence Livermore (CORAL) initiative, the U.S. Department of Energy (DOE) announced the $200 million investment to deliver a next-generation supercomputer, known as Aurora, to the Argonne Leadership Computing Facility (ALCF). When commissioned in 2018, this supercomputer will be open to all scientific users - drawing America's top researchers to Argonne National Laboratory. Additionally, Under Secretary Orr announced $10 million for a high-performance computing R&D programme, DesignForward, led by DOE's Office of Science and National Nuclear Security Administration (NNSA).
The contract is part of the DOE's multimillion dollar initiative to build state-of-the-art supercomputers at Argonne, Lawrence Livermore and Oak Ridge National Laboratories that will be five to seven times more powerful than today's top supercomputers. The CORAL programme will help to advance U.S. leadership in scientific research and maintain its position at the forefront of next-generation exascale computing for years to come.
Intel was selected as the prime contractor and will work with Cray Inc. as the system integrator and manufacturer of these next-generation high-performance computing (HPC) systems for the ALCF. The largest system, to be called Aurora, is based on Intel's HPC scalable system framework and will be a next-generation Cray "Shasta" supercomputer, the successor to the current, industry-leading Cray XC line of supercomputers. The Aurora system will be delivered in 2018 and have a peak performance of 180 petaflops, making it the world's most powerful system currently announced to date. This marks the first selection for a next-generation Cray Shasta supercomputer.
"The ALCF is focused on providing the computational science community with advanced supercomputing systems capable of achieving important breakthroughs in science and engineering, and the Aurora system is a significant continuation of our mission", stated Rick Stevens, Associate Laboratory Director for Argonne National Laboratory. "The scalable architecture of Cray's next-generation Shasta system combined with Intel's advanced processing technologies ensures the Aurora system will be a valuable resource for computational science and engineering."
"The selection of Intel to deliver the Aurora supercomputer is validation of our unique position to lead a new era in HPC", stated Raj Hazra, vice president, Data Center Group and general manager, Technical Computing Group at Intel. "Intel's HPC scalable system framework enables balanced, scalable and efficient systems while extending the ecosystem's decades of software investment to future generations. We look forward to the numerous scientific discoveries and the far-reaching impacts on society that Aurora will enable."
The Aurora system will be 18 times more powerful than its predecessor, Mira, while utilizing only 2.7 times the energy usage. Research goals for the Aurora system include more powerful, efficient and durable batteries and solar panels; improved biofuels and more effective disease control; improving transportation systems and enabling production of more highly efficient and quieter engines; and wind turbine design and placement for improved efficiency and reduced noise.
Contracts valued at more than $200 million are in addition to the recent selections of Intel architecture for the DOE's next-generation Trinity and Cori supercomputers. Aurora will be based on Intel's scalable system framework combining multiple Intel HPC building blocks, including future generations of Intel Xeon Phi processors and the Intel Omni-Path Fabric high-speed interconnect technology, a new non-volatile memory architecture and advanced file system storage using Intel Lustre software.
"Argonne National Laboratory's announcement of the Aurora supercomputer will advance low-carbon energy technologies and our fundamental understanding of the universe, while maintaining United States' global leadership in high performance computing", stated Under Secretary Orr. "This machine - part of the Department of Energy's CORAL initiative - will put the United States one step closer to exascale computing."
Today's $200 million award is the third, and final, supercomputer investment funded as part of the CORAL initiative, a $525 million project announced by Department of Energy Secretary Moniz in November 2014. CORAL was established to leverage supercomputers that will be five to seven times more powerful than today's top supercomputers and help the nation accelerate to next-generation exascale computing. DOE earlier announced a $325 million investment to build state-of-the-art supercomputers at its Oak Ridge and Lawrence Livermore laboratories.
"Few national investments have the potential to demonstrate dramatic progress and capability across many scientific disciplines and domains with real-world benefits", stated Peter Littlewood, Director, Argonne National Laboratory. "Advanced computing is a lever that drives transformational change in science and technology, accelerating discovery and shortening the time for technology to reach market."
Key research goals for the Aurora system, expected to be commissioned in 2018 and to which the entire scientific community will have access, include:
Cray and Intel will provide ALCF with a second system, to be called "Theta", which will be a Cray XC series supercomputer and is expected to be delivered in 2016. Theta will be an early production system for Argonne and will have a peak performance of more than 8 petaflops, while requiring only 1.7 megawatts of power. The Theta system will be powered by Intel Xeon processors and next-generation Intel Xeon Phi processors, code-named Knights Landing, and will be based on the next-generation Cray XC supercomputer. Cray also has options to provide next-generation, high-performance parallel storage systems for both the Theta and Aurora systems.
"We take great pride in building powerful supercomputers for advancing scientific discovery, and through our collaboration with Intel and Argonne, we expect the Aurora system will provide the DOE user community with the most powerful supercomputer we will have ever built", stated Peter Ungaro, president and CEO of Cray. "Aurora also has added significance for us as it will be based on our next-generation Shasta system - the supercomputing heir to our Cray XC series, which is our most successful system ever and a leader in the supercomputing market today. Our Shasta platform will merge the best technologies from our XC line, as well as from our Cray CS line of cluster supercomputers, into a truly adaptive computing framework across both supercomputing and analytics workloads. We are very excited to have been selected by Intel to partner with them to build this unique, ground-breaking system for Argonne."
The next-generation Cray Shasta platform will build upon the advanced supercomputing technologies currently available in the high-end Cray XC and Cray CS cluster supercomputers to provide customers with more flexible, reliable and scalable systems for quickly solving the world's toughest computational challenges. Using best-in-class processors, memory and networks, the Shasta systems are the full embodiment of Cray's Adaptive Supercomputing vision and will be designed to provide Cray customers with a path to exascale computing.
Cray's Shasta platform will feature a next-generation system architecture with the flexibility to support multiple infrastructures and software environments. This will allow Shasta systems to be configured to address a broad range of compute and data-intensive workflows, and the systems will be designed to provide configuration options across the price/performance spectrum to address the different requirements of commercial, government, academic and other user organizations. Shasta's adaptive supercomputing infrastructure is designed to support multiple Intel building blocks, such as future Intel Xeon and Xeon Phi processors and Intel Omni-Path Fabric high speed interconnect technology.
"The Aurora system will represent the pinnacle of supercomputing, providing Argonne with a balanced, power efficient and reliable system based on Intel Architecture that will preserve their existing investments in software applications for years to come", stated Raj Hazra, vice president, Data Center Group and general manager, Technical Computing Group at Intel. "We are excited about this new working relationship with Cray to accelerate the path to exascale and drive the next-generation of scientific discoveries at Argonne National Labs."
A range of positive financial outcomes exist with this contract due to multiple options associated with the various deliverables and funding including a separate R&D contract. Delivery for the initial Theta system is expected in 2016. If exercised as anticipated, Cray expects the larger follow-on Aurora system to be delivered in 2018.
In addition to procuring systems like Aurora, the Office of Science and the National Nuclear Security Administration are making longer-term investments in exascale computing under the DesignForward high-performance computing R&D programme, designed to accelerate the development of next-generation supercomputers. The programme recently awarded $10 million in contracts to AMD, Cray, IBM and Intel Federal, complementing the $25.4 million already invested in the first round of DesignForward. Under this public-private partnership, the four technology firms will work with DOE researchers to study and develop software and hardware technologies aimed at maintaining the USA's lead in scientific computing.