Welcome, Guest. Please login or register.

Sandia National Labs Visualizes Some of the World's Largest Simulations Using NVIDIA Technology

National Laboratories Leverage NVIDIA Quadro Graphics and ParaView Software to Render Extremely Large Datasets in Record Time

Sandia National Labs, Kitware, Inc., and NVIDIA Corporation (Nasdaq: NVDA) today announced a major advance in large-data scientific visualization and rendering for simulations being run on Sandia’s new supercomputer, Red Storm. Sandia’s Red RoSE visualization cluster is being used to render this data, allowing scientists to analyze real world problems with extraordinary precision and fidelity in fields such as national defense and security, aerospace and environmental research.

These results were achieved with ParaView (www.paraview.org), an open source visualization application, running on Red RoSE, one of the world’s most powerful visualization clusters. Developed by Kitware, Inc. (www.kitware.com), ParaView enables parallel graphics processing across a large number of NVIDIA Quadro cards in the Red RoSE visualization cluster. This achievement stems directly from sustained efforts to deliver scalable visualization technologies through strategic investments and developments by the National Nuclear Security Administration’s (NNSA) Advanced Simulation and Computing (ASC) program and the ASC tri-labs, Sandia National Laboratories (SNL), Los Alamos National Laboratory (LANL), and Lawrence Livermore National Laboratory (LLNL).

Visualization is an integral component of ASC and is essential to understanding the massive data produced in simulations being run in the national labs. Such simulation results can be enormous (often terabytes in size) and can provide a unique challenge to the visualization systems tasked with the processing of this data.
The “Red RoSE” cluster, which is part of the infrastructure deployed at Sandia in support of the Red Storm environment, includes 264 visualization nodes (workstations), each comprised of the following:
-- NVIDIA Quadro FX 3400 PCI Express graphics boards
-- Workstations equipped with Dual 3.6 GHz Intel Xeon EM64T processors, and 4GB of RAM
-- InfiniBand 4x HCA interconnect, allowing the data to be processed across the 264 nodes

“The combination of ParaView and NVIDIA high-performance graphics hardware has greatly enhanced our scientific visualization abilities for large data sets. We’re now able to load up, process, and visualize some of the world’s largest datasets at interactive rates,” said Brian Wylie, visualization team leader, Sandia National Labs.

“NVIDIA Quadro commercial off-the-shelf solutions provide the graphics processing power in visualization systems that have traditionally relied on costly proprietary supercomputers,” said Jeff Brown, general manager for NVIDIA professional products, “NVIDIA professional platforms allow our customers like Sandia to deploy solutions that are highly scalable and take advantage of the very latest advances in GPU technologies.”

ParaView performs various operations on the data including calculating derivative scalar fields, t-stripping, cutting, clipping, and glyphing at interactive rates. The aggregate parallel rendering performance is delivered via a high performance image delivery system to the user’s desktop. For typical simulation results, ParaView streams images from a cluster to a desktop at about 15 frames per second. This level of performance is enabled by the latest generation of NVIDIA Quadro graphics hardware which provides fast PCI Express read-back rates. ParaView can deliver rendering performance of over 8 billion polygons per second through Sandia’s parallel rendering library (Ice-T) and Sandia’s image compression system (SQUIRT).

ParaView is built on top of the popular Visualization Toolkit (VTK) and leverages cutting-edge parallel rendering algorithms and leading technology in commodity PC clusters and graphics hardware to interactively visualize some of the world's largest datasets. Contributors and developers currently include: Kitware, LANL, SNL, and the Army Research Laboratory. ParaView is funded by the US Department of Energy ASC program as part of a contract awarded to Kitware, Inc. by a consortium of the three ASC National Labs - LANL, SNL, and LLNL. The goal of the project is to develop scalable parallel processing tools with an emphasis on distributed memory implementations. The project includes parallel algorithms, infrastructure, I/O, support, and display devices. One significant feature of the contract is that all software developed is to be delivered open source. Hence ParaView is available as an open-source system.
News was posted/written by Godboko, on Tue, 15 November 2005 18:04:11 , and has been read 123 times. It is filled under the following

Related Aritcles

There are no comments attached to this item.

Post a new comment

BBCode Actions : Close Tags

Name : Anonymous



Your Ad Here