informa
/
News

NASA Unveils 128-Screen Visualization System

The "hyperwall-2" consists of a 23-foot-wide, 10-foot-tall liquid crystal display capable of rendering a quarter-billion pixel graphics.

NASA's Mega-Wall Of High-Def TVs

The new 128-screen hyperwall-2 system, at NASA's Ames Research Center in California, can render a quarter-billion pixel graphics.
(click for larger image)

NASA announced it has developed the world's highest resolution visualization and data exploration system, with 128 screens and the size of a billboard.

The new hyperwall-2 system, unveiled at Ames Research Center in California this week, can render a quarter-billion pixel graphics, giving scientists rapid pictures of data sets and simulations that would otherwise take years to analyze, NASA said. Scientist and engineers at NASA's Advanced Supercomputing Division (NAS) developed the system with Colfax International.

It consists of a 23-foot-wide, by 10-foot-tall liquid crystal display, 128 graphics processing units, 1,024 processor cores with 74 teraflops of peak processing power, and a data storage capacity of 475 TB. NASA said the hyperwall-2's graphics processing capacity equals that of 600 video-game consoles and exceeds NAS's original 49-screen, hyperwall system by 100 times. It has a direct, high-speed connection to NAS's supercomputers.

The system will help researchers examine, analyze, and relay results from high-fidelity modeling and simulations meant to improve the safety of spacecraft design and analyze conditions for shuttle re-entry. It will also support studies of earthquakes, climate change, weather, and black hole collisions. NASA said the system will produce video and images in ways that make it easier to grasp information.

"The hyperwall-2 offers a supercomputer-scale environment that is truly up to the task of visualization and exploration of the very large datasets routinely produced by NASA supercomputers and instruments," Bryan Biegel, deputy chief of NAS, said in an announcement. "The system also will be used to get highly detailed information on how NAS supercomputers are operating, enabling staff to quickly and precisely diagnose problems or inefficiencies with the supercomputers or the software running on them."