Software Defined



Courtesy of the Texas Advanced Computing Center and University of Texas at Austin - Center for Agile Technology and the US Department of Energy1

software libraries used for professional rendering and scientific visualization

Improve the visual fidelity, management, and efficiency of visualization solutions with this open source initiative from Intel and its partners. Support big data use on supercomputing clusters without the memory limits and cost of GPU-based solutions and unlock the parallelism already in your system. A preconfigured solution—the Software Defined Visualization (SDVis) Appliance—offers in-situ, postprocessing, and professional rendering of visualization tasks.

Learn More

Introduction to SDVis

Visualization Demo

Software Libraries

Enhance existing applications with rendering libraries for parallel software.

Rendering application engineers use these ray-tracing kernels to improve application performance. The kernels are optimized for photo-realistic rendering on the latest processors from Intel with support for Intel® Streaming SIMD Extensions [4.2] and Intel® Advanced Vector Extensions 512.

Courtesy of Attila Afra, Intel2

This portable ray-tracing engine delivers high-performance, high-fidelity visualization for CPUs on Intel® architecture. The rendering library allows you to create rendering applications for interactive applications. OSPRay builds on top of Embree and the Intel® SPMD Program Compiler (Intel® SPC).

Overview of OSPRay 1.0

Courtesy of Ingo Wald and Carson Brownlee, Intel3

This is a high-performance, scalable, OpenGL*-compatible, software rasterizer in the Mesa open source community project. Use unmodified visualization software to work with datasets when GPU hardware is unavailable or is limiting. This CPU-based product runs on laptops, workstations, and compute nodes in high-performance computing (HPC) systems.

Mesa OpenGL* Library

Courtesy of Silvio Rizzi and Joe Insley, Argonne National Laboratory4

Visit Us at SIGGRAPH

See a flexible and powerful 10 node render-focused cluster solution—utilizing the Embree, OSPRay, and OpenSWR libraries—in the Intel Booth at SIGGRAPH 2017, July 30 – August 3 in Los Angeles, CA.



1. Data from Model for Prediction Across Scales for Oceans (MPAS-Ocean) and Accelerated Climate Modeling for Energy (ACME) at the United States Department of Energy. Visualization from Texas Advanced Computing Center (TACC) and University of Texas at Austin - Center for Agile Technology (UT-CAT), Ocean Vortices.

2. Data is reproduced under the Evermotion Commercial License Agreement. Visualization from Attila Afra, Intel, Mazda.

3. Data from Computational Framework for Launch, Ascent, and Vehicle Aerodynamics (LAVA), courtesy of Mike Barad and Cetin Kiris, NASA Ames. Visualization from Ingo Wald and Carson Brownlee, Intel, Landing Gear.

4. Data courtesy of Salman Habib, Katrin Heitmann, and Hardware/Hybrid Accelerated Cosmology Code (HACC) team at Argonne National Laboratory. Visualization from Silvio Rizzi and Joe Insley, Argonne Leadership Computing Facillity, Dark Matter with VL3 and OpenSWR.