SIGGRAPH Asia 2009 was held in Yokohama, Japan. Experts from around the world attended and presented the latest information in graphics technology between December 16th and 19th. NVIDIA, a gold sponsor of SIGGRAPH ASIA 2009, presented several key papers on CUDA development as part of the "GPU Computing Master Class". Here are the presentations given during the class.
SIGGRAPH 2009 NVIDIA Presentations
For 2009, NVIDIA sponsored or presented at 18 sessions, panels, and demos, including 6 special sessions designed to give you up-to-date information on the latest technical innovations from NVIDIA. Below you will find various presentations and videos that we are making publicly available. If you have any feedback about these links, feel free to discuss them on the developer forums at http://devtalk.nvidia.com.
Advances in GPU-based Image Processing and Computer Vision
As a massively parallel processor, the GPU is well-suited for performing "per-pixel" operations in image processing and computer vision. However, new changes in hardware, software, and algorithm mappings allow entire vision algorithms to be performed solely on GPU. This session summarizes how GPU programming now goes beyond per-pixel mappings and is accelerating image-feature processing and handling, frequency domain processing, graph cuts segmentation, and more.
James Fung, NVIDIA Corporation
3D Vision Technology - Develop, Design, Play in 3D Stereo
A detailed technical overview of implementing a 3D stereo experience in your application. Specifically, this session addresses the use cases you must address to create compelling 3D stereo applications, with an emphasis on some key technology challenges and how many of them are easily addressed by some basic implementation patterns. If you're a 3D graphics developer, you will gain a complete set of knowledge on how to implement outstanding 3D stereo applications. This talk will also include a presentation from CAPCOM on their new Resident Evil 5 game, which was build to completely support stereo 3D effects, including both "depth of field" and "out of field" effects.
Samuel Gateau, NVIDIA Corporation
Masaru Ijuuin, CAPCOM
Creating Immersive Environments With NVIDIA APEX
This talk introduces APEX, which greatly speeds up content development by providing artist-friendly tools as well as a framework that enables turn-key solutions without a lot of engineering effort. Find out about our newly expanded suite of APEX packages (Matrix-like destruction, cinematic smoke, free-flowing clothing, fully destructible forests, interactive leaves and other ground debris, and weapon-impact effects) and see real-world game examples in which artists were quickly able to convert static environments into dynamic ones. Discover how you can enhance your games and provide a truly immersive game experience.
Jean Bordes and Monier Maher, NVIDIA Corporation
Alternative Rendering Pipelines on NVIDIA CUDA
This session shows how NVIDIA CUDA can be used for solving complex graphics problems. Ray tracing and REYES are known for being hungry consumers of computational power, and CUDA is a perfect tool for increasing performance. The speakers present their approaches to implementing ray tracing and REYES on CUDA, and show how to use CUDA features to achieve interactive performance.
Andrei Tatarinov and Alexander Kharlamov, NVIDIA Corporation
Efficient Ray Tracing on NVIDIA GPUs (OptiX)
Learn about a new general programming interface for conducting incredibly fast ray tracing on NVIDIA GPUs using C for CUDA. This new technology is valuable for anyone who wants to build a high-performance tracing renderer (interactive or off-line), accelerate an existing ray-tracing renderer, add ray-trace capabilities to raster renderers, or even perform generic ray-tracing functions.
Steve Parker and Phillip Miller, NVIDIA Corporation
Accelerating Realism With the NVIDIA Scene Graph (SceniX)
The NVIDIA Scene Graph quietly powers many of the world's most demanding real-time design environments. Learn how applications can adopt this technology to greatly increase their real-time performance, interactive scene size, and rendering realism through a highly tuned scene graph, distributed GPU rendering, CgFX, and now even ray tracing.
Holger Kunz and Phillip Miller, NVIDIA Corporation
Multi-Layer, Dual-Resolution Screen-Space Ambient Occlusion
This talk describes a general method for rendering higher-quality screen space ambient occlusion (SSAO) by using depth-peeled layers and an enlarged frustum, as well as a dual-resolution method for improving the performance of any SSAO algorithm.
Efficient Substitutes for Subdivision Surfaces
The complete tutorial page can be found athttp://www.nitianyun.com/sigcourse09.html
This presentation below is the implementation section of the tutorial
The goal of this course is to familiarize attendees with the practical aspects of subdivision surfaces for which we introduce substitutes for increased efficiency in real-time applications. The course starts by highlighting the properties that make SubD modeling attractive and introduces some recent techniques to capture these properties by alternative surface representations with a smaller foot-print. We list and compare the new surface representations and focus on their implementation on current and next-generation GPUs. Among the advantages and disadvantages of each approach, we address crucial practical issues, such as watertight evaluation, creases and corners, and seamless displacement mapping. Finally and most importantly, Valve and Industrial Light Magic will present a few breathtaking practical examples and demonstrate how these advanced techniques have been adopted into their gaming and movie production pipelines.
Tianyun Li and Ignacio Castaño, NVIDIA Corporation