In this video from SC16, Jefferson Amstutz from Intel describes how Software Defined Visualization powers a cosmology application running on Intel Xeon Phi.
“Software Defined Visualization (SDVis) is an open source initiative from Intel and industry collaborators to improve the visual fidelity, performance and efficiency of prominent visualization solutions – with a particular emphasis on supporting the rapidly growing “Big Data” usage on workstations through HPC supercomputing clusters without the memory limitations and cost of GPU based solutions. Existing applications can be enhanced using the high performing parallel software rendering libraries OpenSWR, Embree, and OSPRay. At the Intel HPC Developer Conference, Amstutz provided an introduction to this initiative, its benefits, a brief descriptions of accomplishments in the past year and talk about the changes made to Intel provided libraries in the past year.”
Intel-Supported Software Defined Visualization (SDVis):
- Enables deeper understanding of increasingly larger and complex data
- Runs on general purpose processor-based platforms and compute clusters, eliminating the need for GPU-based visualization hardware
- Empowers the ecosystem through open-source software and multiple collaborations
Absolutely unbelievable, Jeff! Your Grandmother Amstutz applauds your willingness to invest your outstanding abilities and the
effort involved in this innovative demonstration! So proud of you, dear grandson!