Anon’s notes on http:// nvidia.com/object/exascale-supercomputing.html

NEXT-GENERATION U.S. DEPARTMENT OF ENERGY SUPERCOMPUTERS WILL BE GPU-ACCELERATED

As GPUs advance in both performance and energy efficiency, accelerated systems have become the new standard for high performance computing (HPC) and a critical ingredient in the pursuit of exascale computing. That's why the U.S. Department of Energy is building two GPU-accelerated supercomputers - as part of the CORAL - that will move the world closer to exascale.

Oak Ridge National Laboratory's "Summit" and Lawrence Livermore National Laboratory's "Sierra" will use NVIDIA® Tesla® GPU accelerators and NVIDIA NVLink™ high-speed interconnect technology on next-generation IBM POWER servers. When they go live, each will be significantly faster than today's fastest supercomputer, achieving between 100 to 300 petaflops of peak performance.

NVIDIA NVLINK HIGH-SPEED INTERCONNECT

Maximum Throughput for Superior Application Performance

NVIDIA® NVLink™ is a high-bandwidth, energy-efficient interconnect that enables ultra-fast communication between the CPU and GPU, and between GPUs. The technology allows data sharing at rates 5 to 10 times faster than the traditional PCIe Gen3 interconnect, resulting in dramatic speed-ups in application performance and creating a new breed of high-density, flexible servers for accelerated computing Download the whitepaper for more details on NVLink.

Designed to meet the challenges of exascale computing, NVLink is a fundamental ingredient of the U.S. Department of Energy's next-generation supercomputers. One such system—"Titan" at Oak Ridge National Laboratory—is currently the fastest supercomputer in the US*. Read more about these supercomputers by downloading the technology brief. Also, for Deep Learning applications, Tesla GPUs can accelerate you time to results from months to days. Read more about deep learning on Tesla platforms here.

• As of April, 2017. Source: www. top500.org/lists/2016/11/

Terminology:

GPU: Graphics processing unit - a specialized electronic circuit designed to rapidly manipulate and alter memory to accelerate the creation of images in a frame buffer intended for output to a display device. GPUs are used in embedded systems, mobile phones, personal computers, workstations, and game consoles. More at Wikipedia

Tesla: Nvidia Tesla is Nvidia's brand name for their products targeting stream processing or general-purpose GPU. Products use GPUs from the G80 series onward. The underlying Tesla microarchitecture of the GPUs and the Tesla product line are named after pioneering electrical engineer Nikola Tesla.