Mercury Computer Systems, NVIDIA Advance Data Computation for Oil Industry

14 Nov, 2007

Mercury's Open Inventor 3D visualization toolkit to integrate with NVIDIA's Tesla GPU HPC solution to deliver advanced parallel-computing capabilities for E&P applications.

Mercury Computer Systems is combining its Open Inventor 3D development toolkit for the simulation and analysis of exploration and production (E&P) workflows with NVIDIA's Tesla GPU compute power. The company states that the partnership will result in enhanced computing capability while facilitating on-the-fly interoperability between 3D visualizations and computations.

"3D visualization has revolutionized the understanding of seismic data, thanks to the performance provided by the GPU," said Jean Bernard Cazeaux, vice-president of the Visualization Sciences Group at Mercury Computer Systems. "GPUs allow much more than visualization -- they provide amazing computing capabilities for interactive applications. Mercury has facilitated the interoperability of Open Inventor with NVIDIA's CUDA language to provide application developers with a unique, integrated solution."

Open Inventor comes with VolumeViz LDM, a volume visualization technology for large seismic datasets (pre- and poststacks) and ReservoirViz LDM, advanced visualization components for reservoir modeling and simulation. The product also includes support for OpenGL shaders. The combination of CUDA/Tesla with Open Inventor reportedly will provide application developers with:

dynamic or static computation capabilities on seismic data with simultaneous 3D visualization feedback
seamless integration of optimized CUDA computing codes into Open Inventor- and VolumeViz LDM-based applications
ability to leverage latest NVIDIA software and hardware solutions (CUDA SDK, Tesla, and Quadro FX boards) to perform intensive parallel computation on the workstation

"Geophysicists are used to taking advantage of their 3D workstation to extract the best data from their seismic volumes during QC and interpretation. Some tasks require access to a supercomputing datacenter to run very complex algorithms on large and complex data volumes. By providing the kind of processing power normally found in a supercomputer on a standard workstation, NVIDIA and Mercury are enabling a fundamental transition in the way you define your workflow," said Andy Keane, general manager of the GPU Computing business at NVIDIA. "By dramatically reducing computation times locally, in some cases from days to hours, NVIDIA Tesla represents the single most significant disruption the high-performance computing industry has ever seen."