Intel® Parallel Computing Center at SCI Institute, University of Utah

SCI Institute logoUniversity of Utah logo

Principal Investigator:

Aaron Knoll

Aaron Knoll is a Research Scientist at the SCI Institute, University of Utah. He completed his Ph.D. at the University of Utah in 2009 on optimized ray tracing methods for implicit surfaces. He has authored papers on efficient CPU and GPU visualization published in leading journals and conferences including IEEE Visualization, ACM Siggraph, IEEE TVCG, and Computer Graphics Forum. At Argonne National Laboratory, he developed Nanovol, a GPU volume raycasting application for computational chemistry problems. At TACC, he supported users of Paraview and VisIt, and developed new rendering technology for the Intel® Xeon Phi™ architecture on the Stampede supercomputer. At SCI, he is continuing his collaboration with Argonne National Laboratory on visualizing large materials data, and pursuing new research in unstructured and particle visualization in the Visus framework, as part of the DOE Scalable Data Management, Analysis and Visualization (SDAV) initiative.


The goal of the Intel® Parallel Computing Center(s) (Intel® PCC) for Scientific Rendering is to research and develop visualization techniques for Intel hardware, finding solutions for large scale and in situ visualization that would not be effective with existing methods. In partnership with the Software Defined Visualization (SDVis) initiative at the Intel Technical Computing Group (TCG), SCI proposes to:

  • Develop techniques for rendering of large unstructured data, particularly of particle data, in the Intel OSPRay framework, including support for unstructured VTK data.
  • Port, optimize and extend the University of Utah’s Visus technology to be the premier parallel in situ visualization tool based on Intel® architecture, supporting streaming LOD visualization of petascale HPC simulations.
  • Port, optimize and extend the University of Utah’s SCIRun software to support CPU visualization, providing a virtual workbench for larger biomechanical simulations on SMP workstation hardware.
  • Increase awareness of Intel’s visualization framework as well as OSPRay and Embree technologies.

Our effort will showcase Intel’s commitment to scientific advancement, and the performance of its CPU and Xeon Phi hardware in GPU-dominated applications of graphics and visualization.


Related Websites:

For more complete information about compiler optimizations, see our Optimization Notice.