You can sign up the October Developer Dinner Event on Parallel Computing here.

Parallel Computing and GPU-based Applications

The topic for the event consists of three parts on parallel computing: an overview of Microsoft’s parallel computing platform and programming tools and facilities, followed by NVidia’s discussion of GPU computing via Visual Studio, and MotionDSP’s discussion on GPU-based video processing capabilities.

Microsoft and Parallel Computing

An overview of the parallel programming tools and facilities available from Microsoft including our activities in the area of GPU computing.  This will include an update on some interesting tools and trends we see in the ecosystem.

David Rich Director, Technical Computing Strategic Business Development. David works in Microsoft’s Technical Computing group with a focus on partnerships and new business opportunities. Previous to Microsoft he was VP marketing at Interactive Supercomputing (acquired by Microsoft) and director of marketing at AMD where he started and ran AMD’s HPC and high-end embedded businesses.  He also served as president of the Hypertransport Consortium.  His experience includes VP of biz dev and marketing assignments at Fujitsu, and startups API Networks and Dolphin Interconnect.  He started and managed the TotalView parallel debugger business and in the early days worked at BBN on Butterfly parallel computers and at Apollo. His only job writing code was Z80 assembler for an SNA (3274) protocol converter.  He has a degree in computer science from Brown University.

Parallel Nsight: GPU Computing via Visual Studio

A brief introduction to how Microsoft and NVIDIA are delivering the potential of GPU accelerated computing integrated in Visual Studio. Highlights will include video demos of the CUDA debugger and powerful Profiling tools. 

John Ashley is NVIDIA’s New York City based Senior Solutions Architect covering Financial Services for North America, Europe, and Africa.  NVIDIA Solution Architects partner at no charge with customers, consultants, Universities and OEMs to educate, evangelize, encourage and support GPU (Graphical Processing Unit) Computing projects. 

GPU-Based Video Processing in Full Motion Video Applications

MotionDSP makes advanced video processing software. MotionDSP’s “Ikena” products improve the quality of video captured by a wide range of sources--from surveillance cameras to unmanned aerial vehicles (UAVs) in law enforcement, homeland security, and military applications. MotionDSP’s technology supports operational deployments within the Department of Defense and National Intelligence communities, and its customers include some of the world’s leading video forensic labs, such as the London Metropolitan Police and the US Secret Service. Results from Ikena have been used successfully in court cases in both the US and UK. 
Recent proliferation of various video sources and the ease of digital distribution has resulted in an exponential rise in the amount of video data that needs to be processed, both during the mission (in real time), and post-mission (offline). Historically, custom application-specific integrated circuits (ASICs) and field-programmable gate arrays (FPGAs) have been required for all demanding image processing tasks. Now, graphics processors (GPUs) provide orders of magnitude more computing power at a fraction of the cost.
In this talk, you will hear how MotionDSP uses GPUs to provide previously-unattainable video processing capabilities to full motion video (FMV) applications. The talk will also include a demonstration of the Ikena software and will provide the details on the design and performance challenges of MotionDSP's video technology.

Nikola Bozinovic is Vice President of Engineering at MotionDSP, leading the team that developed MotionDSP's real-time video processing technology. He has more than ten years of experience in video processing, which includes positions at Microsoft Research Asia and Veodia. He was awarded 2006 EURASIP award for his work on frequency-domain motion analysis. Nikola holds a Ph.D. in Electrical and Computer Engineering from Boston University. For more information about MotionDSP, visit their website.