GPU Computing: Past, Present, and Future

Computer Science and Electrical Engineering
University of Maryland, Baltimore County

GPU Computing: Past, Present, and Future

Dr. David Luebke
Director of Research, NVIDIA Corporation

1:00-2:15pm Friday, 4 February 2011, ITE 227

Modern GPUs have outgrown their graphics heritage in many ways to emerge as the world's most successful parallel computing architecture. The GPUs that consumers buy to play video games provide a level of massively parallel computation in a single chip that was once the preserve of supercomputers. The raw computational horsepower of these chips has expanded their reach well beyond graphics. Today's GPUs not only render video game frames, they also accelerate astrophysics, video transcoding, image processing, protein folding, seismic exploration, computational finance, radioastronomy, heart surgery, self-driving cars – the list goes on and on.

When thinking about the future of GPUs it is important to reflect on the past. How did this peripheral grow into a processing powerhouse found everywhere from medical clinics to radiotelescopes to supercomputers? Why the graphics card and not the modem, or the mouse? Have GPUs really outgrown graphics and will they thus evolve into pure HPC processors? (hint: no)

This talk is intended as a sort of "state of the union" for GPU computing. I'll briefly cover the dual heritage of GPUs, both in terms of supercomputing and the evolution of fixed function graphics pipelines. I'll discuss "computational graphics", the evolution of graphics itself into a general-purpose computational problem, and how that impacts GPU design and GPU computing. Finally I'll describe the important problems and research topics facing GPU computing practitioners and researchers.

David Luebke helped found NVIDIA Research in 2006 after eight years on the faculty of the University of Virginia. Luebke received his Ph.D. under Fred Brooks at the University of North Carolina in 1998. His principal research interests are GPU computing and real-time computer graphics. Luebke's honors include the NVIDIA Distinguished Inventor award, the NSF CAREER and DOE Early Career PI awards, and the ACM Symposium on Interactive 3D Graphics "Test of Time Award". Dr. Luebke has co-authored a book, a SIGGRAPH Electronic Theater piece, a major museum exhibit visited by over 110,000 people, and dozens of papers, articles, chapters, and patents.

Huguens Jean, '03, '11, Ph.D., electrical engineering, and Clifford Muse '11, information systems, returned to Haiti in March 2010, after the devastating January earthquake, to fulfill their grandfather's last request of them. As he was dying of cancer, he asked that at his funeral they celebrate his life and "find the joy." "I had no idea what that meant until we encountered these people in Haiti," said Jean, "These images of life continuing on."

The brothers missed their grandfather's funeral when the earthquake made travel to Port-au-Prince impossible, but they resolved to find a way to honor his memory. The new documentary film Lift Up," co-directed by Jean and UMBC alumnus Philip Knowlton, records their journey back home.

The film will be screened at 8:00pm this on Thursday January 27 in the UMBC Commons Skylight room. Admission is free.

Work by Professor Hillol Kargupta and his students and colleagues was recognize as one of the top ten data mining case studies at the 2010 IEEE International Conference on Data Mining. The case study submission cited was described in

Hillol Kargupta, Kakali Sarkar, Michael Gilligan, Parag Namjoshi, Sai Subhash Paruchuru, Thiraphat Pongsudhiraks, and Robert Gilligan, "MineFleet, A Distributed Vehicle Performance Data Stream Mining System"

The MineFleet software was designed for commercial fleet owners and fleet management companies who want the power of advanced predictive vehicle data mining for dramatically reducing the operating costs. It uses powerful data stream mining algorithms for modeling, benchmarking, and monitoring of vehicle health, emissions, driver behavior, fuel-consumption, and fleet characteristics. Work on Minefleet was done at UMBC and Agnik, a a data analytics company for distributed, mobile, and embedded environments based in Columbia MD. You can read more about the Minefleet system here.