By Prof. Jian Huang, EECS, University of Tennessee
Nowadays, laptops are a necessity for college students. Thanks to this world-changing technology, personal computing has revolutionized the learning environment from Sci Fi to reality over the last 20 years.
At our school’s oneAPI Center of Excellence, we wonder what other possibilities can be enabled by an open innovation-driven technology ecosystem.
For example, today, the digital media landscape influences what we see, think and even how we live. But how do students take advantage of and become a first-class citizen in this fast-changing environment? Imagine a learning environment where students are creators and can own their personal rendering service. Such rendering service would allow students to develop computer-generated animations in a free and/or low cost, reproducible manner. The service would be cloud-based so that it’s available wherever and whenever needed. The best part? It is now a reality and can be shared with anyone at a fraction of the cost, only a few cents per hour to use and available at any time. This opportunity to gain in-depth experience building interactive cloud services, along with the freedom to explore and innovate visualization unhindered by a PC’s compute constraints, opens doors to innovation exponentially.
I’m proud to share that work within our oneAPI Center of Excellence, a collaboration between Intel and the University of Tennessee, has made a student-created rendering service a reality. Typically, high-end visualization is heavy-weight, cumbersome to deploy, and expensive to use. The Center’s objective was to build high-end visualization as a service that can be instantly available and universally accessible to students.
The open innovation enabled by the Intel oneAPI Rendering Toolkit supports a model that simplifies development across multiple types of architectures. It is based on industry standards and designed to deliver uncompromised performance for accelerated compute and remove constraints of proprietary programming models. oneAPI technologies such as ray tracing/rendering and data and media processing accelerate computing across heterogeneous architectures (CPUs, GPUs—including the upcoming Intel® Data Center GPU Max Series, and other accelerators). While being open sourced and portable does lower development and deployment costs, the real enabler is its ability to quickly prototype and elastically scale up, which comes from the oneAPI libraries.
Armed with Intel OSPRay, we coined the term Visualization as a Service (VaaS) and developed a powerful framework from a high-level concept. Our methods create interactive rendering through elastic parallelism directly from Visualization Cloud Instance (VCI), such as Amazon AWS. The VCIs are containerized and easily deployed on HPC computers, on-prem institutional clusters, as well as through cloud service providers (CSPs). This multiple location strategy provides flexibility for app developers and system administrators with an unprecedented amount of freedom to choose software architectures that are optimal for their users. VCIs are scalable and support CPU platforms with or without GPUs or other accelerators to maximize compute. In an agile manner, developer communities can compose visual dashboards that disseminate knowledge and insights broadly with interactive high-end rendering as the core. Click here for a live demo of VaaS.
Before the student-created rendering service was possible, students could not own a cloud service that let them create and customize digital and graphic content freely. This barrier prevented them from learning powerful new skillsets that are so far restricted to the circles of seasoned professionals in technology companies. For students, broadening their access to such skillsets and technological capabilities that already exceed the industry norm provides them with exciting new opportunities and enables them to contribute to the industry.
Everything we do at the oneAPI Center of Excellence is open source. Anyone can start with our GitHub project: Set up a cloud-based rendering service, which teaches college students (and even motivated high school students) interested in interactivity how to build their own Rendering as a Service (RaaS). In a matter of 15 to 20 hours, a student can learn the process and technical steps to complete a RaaS service of their own on Amazon AWS. This service can hit a computing budget as low as 4 cents/hour yet delivers the capabilities and tools for high-end rendering that was not available to students before.
While this simple student project is not yet changing the world of learning, it does provide young, creative minds access to a fun project and a final product worth showing off forever.
Ongoing work is still taking place to make the open-source project even better. Please star this project:github.com/seelabutk/substrate-ospray-studio to stay informed of our progress.