Sandia LabNews

DOE to deploy Arm-based supercomputer prototype at Sandia

Image of arm_logos_600

Microprocessors designed by Arm are ubiquitous in automobile electronics, cellphones and other embedded applications, but until recently they have not provided the performance necessary to make them practical for high-performance computing.

Astra — one of the first supercomputers to use processors based on the Arm architecture in a large-scale high-performance computing platform — is expected to be deployed at Sandia later this summer.

The DOE’s NNSA announced that Astra, the first of a potential series of advanced architecture prototype platforms, will be deployed as part of its Vanguard program. Vanguard will evaluate the feasibility of emerging high-performance computing architectures as production platforms to support NNSA’s mission to maintain and enhance the safety, security and effectiveness of the U.S. nuclear stockpile.

“One of the important questions Astra will help us answer is how well does the peak performance of this architecture translate into real performance for mission applications,” says Mark Anderson, program director for NNSA’s Advanced Simulation and Computing program, which funds Astra.

Astra will be based on the recently announced Cavium Inc. ThunderX2 64-bit Arm-v8 microprocessor. The platform will consist of 2,592 compute nodes, of which each is 28-core, dual-socket. Its theoretical peak will be more than 2.3 petaflops, equivalent to 2.3 quadrillion floating-point operations (FLOPS), or calculations, per second. While being the fastest machine is not one of the goals of Astra or of the Vanguard program in general, a single Astra node is roughly one hundred times faster than a modern Arm-based cellphone.

A first step for Vanguard

Scott Collis, director of Sandia’s Center for Computing Research, says, “Emerging architectures come with many challenges. Since the NNSA has not previously deployed high-performance computing platforms based on Arm processors, there are gaps in the software that must be addressed before considering this technology for future platforms much larger in scale than Astra.”

“Arm has been deeply engaged with Sandia National Laboratories working to comprehend and deliver on the needs of the high-performance computing community. We are eager to support the Vanguard program as a key milestone deployment for Arm and our partners, delivering on a shared vision to spur innovation in this critical domain.”

— Drew Henry, senior vice president and general manager of Arm’s infrastructure business line

As part of a multilab partnership, researchers anticipate continually improving Astra and future platforms.

“Sandia researchers partnering with counterparts at Los Alamos and Lawrence Livermore national laboratories expect to develop an improved software-and-tools environment that will enable mission codes to make increasingly effective use of Astra as well as future leadership-class platforms,” says Ken Alvin, senior manager for extreme scale computing. “The Vanguard program is designed to allow the NNSA to take prudent risks in exploring emerging technologies and broadening our future computing options.”

Astra will be installed at Sandia in an expanded part of the building that originally housed the innovative Red Storm supercomputer.

The Astra platform will be deployed in partnership with Westwind Computer Products Inc. and Hewlett Packard Enterprise.

“Astra, like Red Storm, will require a very intimate collaboration between Sandia and commercial partners,” says James Laros, Vanguard program lead. “In this case, all three NNSA defense labs will work closely with Westwind, HP Enterprise, Arm, Cavium and the wider high-performance computing community to achieve a successful outcome of this project.”

Astra takes its name from the Latin phrase “per aspera ad astra,” or “through difficulties to the stars.”