$70 Million Contract for Cray for HPC at National Energy Research Scientific Computing Center

Cray Inc. has been awarded a $70 million contract to provide its XC supercomputer to the US Department of Energy’s (DOE) National Energy Research Scientific Computing Center (NERSC).

Named ‘Cori’ after bio-chemist and Nobel Laureate Gerty Cori, NERSC’s next-generation Cray XC supercomputer will provide its vast user community with an advanced supercomputing resource to support the Center’s large production workload. NERSC supports more than 5,000 scientists annually on more than 700 projects in the areas of climate modeling, biology, environmental sciences, combustion, materials science, chemistry, geosciences, fusion energy, astrophysics, nuclear and high-energy physics, and other disciplines, along with scientific visualization of massive data sets.

“We are excited to continue our partnership with Cray,” said Sudip Dosanjh, director, NERSC at Lawrence Berkeley National Laboratory. “Cori will provide a significant increase in capability for our users and will provide a platform for transitioning our very broad user community to many core architectures. We will collaborate with Cray to ensure that Cori meets the computational and data needs of DOE’s science community.” NERSC is also home to a Cray XC30 supercomputer named “Edison” and a Cray XE6 supercomputer named “Hopper.“

The new Cori system is expected to deliver 10 times the sustained computing capability of the Hopper supercomputer.

“We are proud of the history we’ve built with NERSC, and we are honored that the Center, along with the Department of Energy’s Office of Science, has once again turned to Cray to support the future computational needs of their large user community,” said Peter Ungaro, president and CEO, Cray. “This is a significant contract for our company as it demonstrates that our roadmap for the Cray XC family will continue to lead the industry well into the future. The researchers and scientists at NERSC are focused on solving a wide range of challenging problems that demand high levels of performance and reliability across a broad spectrum of scientific applications. We look forward to working with NERSC and putting our future technologies to the test as part of the Department of Energy’s scientific research and discovery program.”

The Cori system to be installed at NERSC will be a future-generation version of the Cray XC30 supercomputer and will feature the next-generation of the Intel Xeon Phi processor code-named ‘Knights Landing.’

“We are thrilled to work with Cray in bringing the next generation of highly parallel supercomputers to market based on the Intel Xeon Phi processor – codenamed Knights Landing,” said Charles Wuischpard, Intel Corp.‘s VP, data center group and GM, workstation and HPC. “Working closely with Cray, we will deploy the Many Integrated Core (MIC) architecture on the next generation Cray XC supercomputer, delivering over 3 teraflops of performance per single socket node to power a wide set of applications and taking an important and viable step towards Exascale.“

Previously code-named ‘Cascade,’ the XC30 series of supercomputers are engineered to meet the performance challenges of demanding HPC users. Special features of the XC30 and XC30-AC supercomputers include: the Aries system interconnect; a Dragonfly network topology that frees applications from locality constraints; innovative cooling systems to lower customers’ TCO; the next-generation of the scalable, performance Cray Linux environment supporting a range of applications; Cray’s HPC optimized programming environment, and the ability to handle a variety of processor types.

Consisting of products and services, the multi-year contract is valued at more than $70 million, and the system is expected to be delivered in 2016. As part of this contract, the next-generation Cray XC supercomputer will also include a 400Gb/s, 28PB Cray Lustre file system storage solution. Additionally, NERSC has the option to purchase solid state storage integrated in the XC supercomputer for high performance burst I/O.