Press Release
As a global leader in supercomputing, Cray provides highly advanced systems and solutions and world-class service.

Contact us

Press Release

Printer Friendly Version View printer-friendly version << Back

U.S. Department of Energy and Cray to Deliver Record-Setting Frontier Supercomputer at ORNL

Exascale system expected to be world's  most powerful computer for science and innovation

OAK RIDGE, TN - The U.S. Department of Energy today  announced a contract with Cray Inc. to build the Frontier supercomputer at Oak  Ridge National Laboratory, which is anticipated to debut in 2021 as the world's  most powerful computer with a performance of greater than 1.5 exaflops.

Scheduled for delivery in 2021, Frontier will accelerate  innovation in science and technology and maintain U.S. leadership in  high-performance computing and artificial intelligence. The total contract  award is valued at more than $600 million for the system and technology  development.  The system  will be based on Cray's new Shasta architecture and Slingshot interconnect and  will feature high-performance AMD EPYC CPU and AMD Radeon Instinct GPU  technology.

"Frontier's record-breaking performance will ensure our country's  ability to lead the world in science that improves the lives and economic  prosperity of all Americans and the entire world," said U.S. Secretary of  Energy Rick Perry. "Frontier will accelerate innovation in AI by giving  American researchers world-class data and computing resources to ensure the  next great inventions are made in the United States."

By solving calculations up to 50 times faster than today's top  supercomputers —exceeding a quintillion, or 1018,  calculations per second—Frontier will enable researchers to deliver  breakthroughs in scientific discovery, energy assurance, economic  competitiveness, and national security. As a second-generation AI system –  following the world-leading Summit system deployed at ORNL in 2018 – Frontier  will provide new capabilities for deep learning, machine learning and data  analytics for applications ranging from manufacturing to human health.
  Since 2005, Oak Ridge National Laboratory has deployed Jaguar, Titan, and  Summit, each the world's fastest computer in its time. The combination of  traditional processors with graphics processing units to accelerate the  performance of leadership-class scientific supercomputers is an approach  pioneered by ORNL and its partners and successfully demonstrated through ORNL's  No.1 ranked Titan and Summit supercomputers.  

"ORNL's vision is to sustain the nation's preeminence in science  and technology by developing and deploying leadership computing for research  and innovation at an unprecedented scale," said ORNL Director Thomas Zacharia.  "Frontier follows the well-established computing path charted by ORNL and its  partners that will provide the research community with an exascale system ready  for science on day one."

Researchers with DOE's Exascale Computing Project are developing  exascale scientific applications today on ORNL's 200-petaflop Summit system and  will seamlessly transition their scientific applications to Frontier in 2021. 

Frontier will offer best-in-class traditional scientific modeling  and simulation capabilities while also leading the world in artificial  intelligence and data analytics. Closely integrating artificial  intelligence with data analytics and modeling and simulation will drastically  reduce the time to discovery by automatically recognizing patterns in data and  guiding simulations beyond the limits of traditional approaches.

"We are honored to be part of this historic moment as we embark on  supporting extreme-scale scientific endeavors to deliver the next U.S. exascale  supercomputer to the Department of Energy and ORNL," said Peter Ungaro,  president and CEO of Cray. "Frontier will incorporate foundational new  technologies from Cray and AMD that will enable the new exascale era --  characterized by data-intensive workloads and the convergence of modeling,  simulation, analytics, and AI for scientific discovery, engineering and digital  transformation."

Frontier will incorporate several novel technologies co-designed  specifically to deliver a balanced scientific capability for the user  community. The system will be composed of more than 100 Cray Shasta cabinets  with high density compute blades powered by HPC and AI- optimized AMD EPYC  processors and Radeon Instinct GPU accelerators purpose-built for the needs of  exascale computing. The new accelerator-centric compute blades will support a  4:1 GPU to CPU ratio with high speed AMD Infinity Fabric links and coherent  memory between them within the node. Each node will have one Cray Slingshot  interconnect network port for every GPU with streamlined communication between  the GPUs and network to enable  optimal performance for high-performance computing and AI workloads at  exascale.

To make this performance seamless to consume by developers, Cray  and AMD are co-designing and developing enhanced GPU programming tools  optimized for performance, productivity and portability. This will include new  capabilities in the Cray Programming Environment and AMD's ROCm open compute  platform that will be integrated together into the Cray Shasta software stack  for Frontier.

"AMD is proud to be working with Cray, Oak Ridge National  Laboratory and the Department of Energy to push the boundaries of high  performance computing with Frontier," said Lisa Su, AMD president and CEO.  "Today's announcement represents the power of collaboration between private  industry and public research institutions to deliver groundbreaking innovations  that scientists can use to solve some of the world's biggest problems."

ORNL's Center for Accelerated Application Readiness is now  accepting proposals from scientists to prepare their codes to run on Frontier. Visit the Frontier website to learn more about what researchers plan to accomplish in  these and other scientific fields.

Frontier will be part of the Oak Ridge Leadership Computing  Facility, a DOE Office of Science User Facility. For more  information, please visit