36

 

*STTR Project:  Chromium RenderServer:  Remote Rendering and Visualization Servers Using Distributed Memory Parallel Clusters—Tungsten Graphics, 416 Good Hill Road, Woodbury, CT  06798; 970-780-4272, www.tungstengraphics.com

Mr. Brian Paul, Principal Investigator, brialpaul@tungtengraphics.com

Mr. Jens Owen, Business Official, jens@tungstengraphics.com

DOE Grant No. DE-FG02-04ER86198

Amount:  $750,000

 

Research Institution

Lawrence Livermore National Laboratory

Livermore, CA 

                                                           

While computational capacity at DOE’s flagship computer centers continues to increase, there is a corresponding increase in need for interactive data analysis.  Although researchers are capable of generating unprecedented amounts of data, they also need access to facilities and software that allow them to gain insight into the scientific phenomena hidden in their simulation or experimental data.  However, most such researchers only have network-based access to the centrally located computer centers, yet the amount of data to be analyzed is so large as to preclude transfer over the network to a local computer system.  This project will define and implement a platform-neutral software infrastructure that provides the ability to perform remote scientific visualization on high performance, interactive, parallel platforms, and transmit the resulting imagery to a remote user or to a remote team of geographically distributed users.  Phase I developed a prototype architecture that allows a remote user to view results generated by a remotely-located graphics or visualization application running on a shared- or distributed-memory computer system.  Phase II will develop a robust, full-featured Chromium Renderserver that delivers high-resolution visualization imagery to one or more remote users from distributed- or shared-memory parallel computer systems.

 

Commercial Applications And Other Benefits as described by the awardee:  The remote delivery of visualization results from scalable rendering platforms would meet a critical need that spans all computational science projects that generate data at centrally located facilities, but need to perform visual data analysis from remote locations.  Commercialization opportunitis include the provision of long term consulting, development, porting and maintenance agreements.