AMD to Power Exascale Cray System at ORNL

Print Friendly, PDF & Email

Energy Secretary Rick Perry introduces the Frontier Supercomputer.

Today AMD announced a new exascale-class supercomputer to be delivered to ORNL in 2021. Built by Cray, the “Frontier” system is expected to deliver more than 1.5 exaFLOPS of processing performance on AMD CPU and GPU processors to accelerate advanced research programs addressing the most complex compute problems.

Today marks a significant milestone for AMD and its high-performance CPU and GPU compute technologies as we take the stage with Cray and Oak Ridge National Laboratory to set a high bar for exascale computing in 2021,” said Lisa Su, president and CEO, AMD. “Frontier represents the latest achievement on a long list of technology innovation AMD has contributed to the Department of Energy exascale programs, including participation in the FastForward, Design Forward and Path Forward programs.

Researchers at ORNL will use the vast computing resource provided by Frontier to simulate, model and advance understanding of the interactions underlying the science of weather, sub-atomic molecular structures, genomics, and physics, to name a few.

The thing that’s really attractive about Frontier is the powerful nodes,” said John Turner from ORNL. “Having fewer powerful nodes with a very tightly integrated set of CPUs and GPUs at the node-level gives us the ability to distribute hundreds or thousands of microstructure and property calculations on one or a few nodes across the machine. With Frontier, we’re going to be able to predict the microstructure and properties of an additively manufactured part at much higher fidelity and in many more locations within a part than we are able to even with the world’s current fastest supercomputers.”

With this project, the best of AMD hardware and software will come together toward the singular vision of solving the toughest challenges in the world today.”AMD will deliver high performance innovation to the“Frontier” system in the Cray Shasta system architecture, including:

  • HPC and AI-optimized custom AMD EPYC CPU and purpose-built Radeon Instinct GPU processors
  • High Bandwidth Memory (HBM) in support of the GPU
  • Tightly integrated 1:4 CPU to GPU ratio
  • Connected by a custom coherent Infinity Fabric
  • Supported by an enhanced, open ROCm programming environment for AMD CPUs and GPUs

We at Cray are humbled and honored to have been selected by the US Department of Energy to deliver the Frontier system to Oak Ridge National Laboratory in partnership with AMD,” said Steve Scott, senior vice president and CTO at Cray. “Cray Shasta supercomputers will deliver the fusion of capabilities needed to power the next era of advanced computing. The combination of a flexible compute infrastructure, scalable HPC and AI software, and the intelligent Slingshot system interconnect will enable Cray customers to undertake a new age of science, discovery and innovation at any scale.”

“Frontier” joins the Department of Energy CORAL-2 initiative to support the research, development, and procurement of two exascale computing systems for the DOE. This program is the result of years of ongoing collaboration between public institutions and private industry which sought vendor solutions that would maximize the energy efficiency of systems while improving the performance, productivity, and reliability for key U.S. DOE applications.

Download the Frontier Specification Sheet (PDF)

Sign up for our insideHPC Newsletter