Illinois Researchers Discuss Multicore Applications at Microsoft Workshop

REDMOND, WA — Microsoft brought together researchers from UPCRC Illinois, UC Berkeley, Intel, and Microsoft to discuss current and future applications that would benefit from multicore parallelism and hasten the adoption of multicore machines. A broad range of novel, interesting, and computationally challenging applications — whether or not they currently take advantage of multicore computers &mdash were presented in four key areas: Visual Computing, Social Interaction, Speech and Audio, and Human-Machine Interaction.

Session I: Visual Computing I

• Parallel Computer Vision Algorithms — Rick Szeliski, Microsoft Research
• Video Event Detection/Hand Tracking — Dennis Lin/Mert Dickman, UPCRC Illinois
• Robust Face Recognition — Andrew Wagner, UPCRC Illinois

Session II: Visual Computing II
NOTE: video/slides are not available for this session

• Medical Imaging and Parallel Computing — Dubey Pradeep, Intel Research
• Image-Based Rendering — Sanjay Patel, UPCRC Illinois
• Advanced MR Image Reconstruction using GPU-level Parallelism — Justin Halder, UPCRC Illinois
• Content Based Image Retrieval — Bryan Catanzaro, UC Berkeley

Session III: Social Interaction

• Intel's Immersive Connected Experience Research — Jim Held, Intel Research
• Telepresense — Philip Chou, Microsoft Research
• Parallelizing Speech Recognition and Making it Better — Nelson Morgan, UC Berkeley

Session IV: Speech and Audio

• The Breadth of Applications for Music — Eric Battenberg, UC Berkeley
• Telepresense — Philip Chou, Microsoft Research

Session V: Human-Machine Interaction I

• Personalized Medicine from Medical Imaging and Advanced Computation — Tony Keaveny, UC Berkeley
• Design for User Experience Applications — Gad Scheaffer, Intel Research
• Optimizing Game Architectures with Task-Based Parallelism — Brad Werth, Intel Research
• Parallelizing Machine Learning: Applications to Recommender Systems, Computer Go, and Bioinformatics — David Stern, Microsoft Research

Session VI: Human-Machine Interaction II

• Computational Challenges of Open-World Intelligence — Eric Horvitz, Microsoft Research
• Dynamic Virtual Environments — John Hart, UPCRC Illinois
• Parallelizing the Web Browser — Ras Bodik, UC Berkeley
• Designing and Implementing Secure Web Browsers — Sam King, UPCRC Illinois

Driving the UPCRC Illinois agenda is a human-centric vision of future consumer applications (i.e., “multicore killer apps”). Investigating applications similar to those above reveals new parallel patterns and serves as a testbed for evaluating, refining, and ultimately proving UPCRC Illinois ideas on multicore programming.

Video and slides of most presentations are available by clicking on the session titles above.

About the Universal Parallel Computing Research Center The Universal Parallel Computing Research Center (UPCRC) at the University of Illinois is a joint research endeavor of the Department of Computer Science, the Coordinated Science Laboratory, the Department of Electrical and Computer Engineering, and corporate partners Microsoft and Intel. The center builds on a history of Illinois innovation in parallel computing that spans four decades. UPCRC is also one of many Parallel@Illinois efforts currently invested in pioneering and promoting parallel computing research and education.