The ddiLab at Northern Illinois University has deployed the Atipa Polaris High Performance Computing and Visualization (HPCV) platform. The platform is a turnkey Intel® Select Solution for Professional Visualization for scientific simulation and visualization.
The Polaris Select HPCV platform leverages the Intel® Rendering Framework with optimized ray-tracing algorithms to use Intel® Xeon® Scalable Processors instead of graphics processors for high-fidelity rendering and visualization. MPI-enabled rendering extends the scalability of visualization workloads beyond a single node and enables the use of the aggregated memory of all cluster nodes to visualize data sets of hundreds of gigabytes to terabytes. This allows the ddiLab students to work locally on a resource that is similar to some of the largest compute systems in the world.
This resource will be used to teach classes in both the School of Art and the Department of Computer Science. It will used actively by the student researchers within the ddiLab for work at both the undergraduate and graduate level in the areas of scientific visualization and high performance computing.
hopcroft.cs.niu.edu: a platform for simulation and analysis
This effort is supported by generous donation of hardware from Intel and Atipa Technologies, with funding support from the Office of the Vice President for Research and Innovation Partnerships at Northern Illinois University and the National Science Foundation (ACI-1532133).
This work uses an optimized version of Visual Molecular Dynamics (VMD) with Intel OSPRay to render the output of a Large-scale Atomic/Molecular Massively Parallel Simulator (LAMMPS) simulation computed at the Argonne Leadership Computing Facility. The gold spheres are Silicene, which is a monolayer silicon, on a metal surface of Iridium, shown in blue. Materials such as Silicene have the potential to be integrated into current semiconductor manufacturing technologies and possess properties that may lead to next-generation of improved electronics, including batteries and sensory devices. [Data courtesy: Subramanian Sankaranarayanan, Argonne National Laboratory]
Use of hopcroft in the analysis of data generated at the Argonne Leadership Computing Facility, the ddiLab is using dedicated resources like this to train next generation visualization scientists. Most recently this work was shown in the Intel booth at the ISC High Performance 2019 in Frankfurt, Germany. [Data courtesy: Lars Bildsten and Yan-Fei Jiang, University of California at Santa Barbara]