There is a campuswide interest in multimedia/VR at Syracuse University, involving labs and departments such as the CASE Center, NPAC, School of Information Studies, Multimedia Lab and Advanced Graphics Research Lab. A small scope virtual reality Lab has been started, sponsored by the CASE Center and Chris Gentile from AGE, Inc., who is an SU alumnus and partner in the successful NYS startup focused on low-end broad-market consumer VR products. New planned collaborations with the corporate sponsors include joint projects with SimGraphics Engineering, Inc., a California-based company developing high-quality graphics software for simulation, animation, and virtual engineering, and with virtual reality, Inc., a new East Coast startup interested in developing high-end VR systems with high-performance computing support.
On the base VR research side, there is a planned collaboration with Rome Laboratories [Nilan:91a] aimed at designing the VR-based group decision support for the modern CI systems. The project also involves evaluating MOVIE as a candidate for the high-end VR operating shell. Within the new multidisciplinary Computational Neuroscience Program at Syracuse University, we are also planning to couple some vision and neutral network research issues with the design issues for VR environments such as ``nonencumbered'' machine vision-based interfaces, VR-related perception limits, or neural net-based predictive tracking techniques for fast VR rendering.
Multimedia is a discipline closely associated with VR and strongly represented at Syracuse University by the Multimedia Lab within the CASE Center and by the Belfer Audio Lab. Some of the multimedia applications are more static and/or text-based than the dynamic three-dimensional VR environments. The borderline between both disciplines is usually referred to as hypermedia navigation-that is, dynamic real-time exploration of multimedia databases. Large, complex databases and associated R&D problems of integration, transmission, data abstraction, and so on, represent the common technology area connecting multimedia and VR projects.
Our interests at NPAC are towards high-performance VR systems, based on parallel computing support. A powerful VR environment could be constructed by combining the computational power and diverse functionality of new parallel systems at NPAC: CM-5, nCUBE2, and DECmpp, connected by fast HIPPI networks. A natural VR task assignment would be: modeller/simulator on CM-5, parallel database server on nCUBE2, and renderer on DECmpp-which basically exhausts all major computational challenges of virtual reality.
The relevance of parallel computing for VR is both obvious and yet largely unexplored within the VR community. The popular computational engine for high-end VR is provided currently by the Silicon Graphics machines and these systems are in fact custom parallel computers. But it remains to be seen if this is the most cost-effective or scalable solution for VR. The most natural testbed setup for exploring various forms of parallelism for VR can be provided by general-purpose systems. The distributed environment described above and based on a heterogeneous collection of general-purpose parallel machines would provide us with truly unique capabilities in the domain of high-end parallel/distributed VR. We intend to develop VR support in MOVIE and to use it as the base infrastructure system for high-end VR at NPAC. We discuss MOVIE's role in the VR area in more detail in the next section.