Overview

There are strong indications that the future of interactive graphics programming is a model more flexible than today's OpenGL/Direct3D pipelines. As such, graphics developers need to have a basic understanding of how to combine emerging parallel programming techniques and more flexible graphics processors with the traditional interactive rendering pipeline. The first half of the course introduces attendees to modern parallel graphics architectures and parallel programming models, and describes current and near-term use of these new capabilities for real-time rendering. The second half of the course looks farther ahead at trends emerging in the academic literature and offline rendering communities as researchers use these many-core parallel architectures to explore future rendering pipelines. Topics include future, and more flexible, rendering pipelines that support true motion blur, depth-of-field, curved surfaces, and complex dynamic lighting. The course concludes with a panel, moderated by the creator of OpenGL Kurt Akeley, on the role of fixed function hardware in future graphics architectures.

The course presenters are experts on advanced rendering, graphics hardware, and parallel computing for graphics from academia and industry, and have presented papers and tutorials on the topic at SIGGRAPH, High Performance Graphics, Supercomputing, IEEE Visualization, and elsewhere.

Course Organizers

Aaron Lefohn, Intel
Mike Houston, AMD

Course Notes

(All slides zip)

Beyond Programmable Shading I

9:00–9:20 Looking Back, Looking Forward, Why and How is Interactive Rendering Changing
Mike Houston, AMD
9:20–9:45 Five Major Challenges in Interactive Rendering
Johan Andersson, DICE
9:45–10:15 Running Code at a Teraflop: How a GPU Shader Core Works
Kayvon Fatahalian, Stanford University (Kayvon's graphics blog)
10:15–10:45 Parallel Programming for Real-Time Graphics
Aaron Lefohn, Intel
10:45–11:15 DirectCompute Use in Real-Time Rendering Products
Chas Boyd, Microsoft
11:15–11:45 Surveying Real-Time Beyond Programmable Shading Rendering Algorithms (PDF with notes)
David Luebke, NVIDIA Research
11:45–12:15 Bending the Graphics Pipeline
Johan Andersson, DICE

Beyond Programmable Shading II

2:00–2:05 Welcome and Re-Introduction
Aaron Lefohn, Intel
2:05–2:35 Keeping Many Cores Busy: Scheduling the Graphics Pipeline
Jonathan Ragan-Kelley, MIT
2:35–3:10 Evolving the Direct3D Pipeline for Real-Time Micropolygon Rendering
Kayvon Fatahalian, Stanford
3:10–3:30 Decoupled Sampling for Real-Time Graphics Pipelines (Full paper at MIT)
Jonathan Ragan-Kelley, MIT
3:30–3:50 Deferred Rendering for Current and Future Rendering Pipelines (Source code) (PDF with notes)
Andrew Lauritzen, Intel
3:50–4:15 PantaRay: A Case Study in GPU Ray-Tracing for Movies
Luca Fascione, WETA and Jacopo Pantaleoni, NVIDIA Research
4:15–4:30 Wrapup: What's Next for Interactive Rendering Research?
Mike Houston, AMD
4:30–5:15 Panel: What Role Will Fixed-Function Hardware Play in Future Graphics Architectures?
Moderator
Kurt Akeley, Microsoft Research
Panelists
Steve Molnar, NVIDIA
David Blythe, Intel
Mike Houston, AMD
Johan Andersson, DICE
Kayvon Fatahalian, Stanford