NEWPORT NEWS, VA. – The fastest computer in Virginia is located at the Department of Energy's Thomas Jefferson National Accelerator Facility. The lab’s newest supercomputer debuted as the 95th fastest in the world on the TOP500 Supercomputer Sites list.
This is Jefferson Lab's first appearance on the TOP500 list. The 30th edition of the list was released Monday at SC07, the international conference on high performance computing, networking, storage and analysis, in Reno.
The lab’s supercomputer runs powerful computer simulations to shed light on how one of the basic forces of nature, the strong force, builds protons, neutrons and other particles from the basic building blocks of matter: quarks and gluons.
The supercomputer, dubbed 7N, is a cluster computer composed of individual units wired together to function as one. It is built of 396 nodes, or individual computers, each containing two Advanced Micro Devices (AMD) quad-core 1.9 GHz processors with four Gigabytes (GB) of onboard random access memory (RAM). The nodes are connected via a Double Data Rate Infiniband network, which is capable of transferring 20 Gigabits of data per second (Gb/s) between the nodes.
"With our cluster computers here at Jefferson Lab, we have always taken the cost-effective way,” says Jie Chen, a Jefferson Lab Senior Computer Scientist and a lead researcher in setting up and testing the supercomputer. "This listing puts Jefferson Lab on the map in the scientific computing community."
The cluster was installed in Jefferson Lab’s Computer Center by the members of the lab’s High-Performance Computing Group in late summer. Once installed, it more than tripled the lab’s high-end computing capabilities.
"We're proud that the Department of Energy has provided this incredible resource here at Jefferson Lab for our research into the fundamental structure of matter. Hampton Roads is now home to a supercomputer that's among the world leaders in scientific computing," says Jefferson Lab Chief Information Officer Roy Whitney.
The cluster qualified for the list by running the Linpack Benchmark, a calculation that’s used by the TOP500 as a yardstick for supercomputer performance. The cluster ultimately clocked out at 13460 Gigaflops running solidly on 380 nodes.