| Date | Detail |
| Now Closed | Call for Sessions |
| January 9, 2012 | Notification of Acceptance for Sessions |
Call for Sessions for GTC U.S. is now closed. If you have any questions, please contact us at gtc-submissions@nvidia.com.
To view last year's presentations check out the GTC 2010 Archives.
Important Information on How to Write a Great Session Description
The first sentence should explain what the session registrant can expect to learn, see, hear, or do. Get right to the point, avoid background your audience already knows (e.g., "Originally designed as graphics accelerators, GPUs have evolved into powerful parallel processors capable of accelerating many compute-intensive applications.") Subsequent sentences should offer more details about what will be covered and why the reader should attend. In general, go for clarity over cleverness.
The description should begin with an action word such as:
Length:
Good Examples from GTC 2010:
Faster, Cheaper, Better – Hybridization of Linear Algebra for GPUs
Learn how to develop faster, cheaper and better linear algebra software for GPUs through a hybridization methodology that is built on (1) Representing linear algebra algorithms as directed acyclic graphs where nodes correspond to tasks and edges to dependencies among them, and (2) Scheduling the execution of the tasks over hybrid architectures of GPUs and multicore. Examples will be given using MAGMA, a new generation of linear algebra libraries that extends the sequential LAPACK-style algorithms to the highly parallel GPU and multicore heterogeneous architectures.
Analysis-Driven Performance Optimization
The goal of this session is to demystify performance optimization by transforming it into an analysis-driven process. There are three fundamental limiters to kernel performance: instruction throughput, memory throughput, and latency. In this session we will describe: how to use profiling tools and source code instrumentation to assess the significance of each limiter; what optimizations to apply for each limiter; how to determine when hardware limits are reached. Concepts will be illustrated with some examples and are equally applicable to both CUDA and OpenCL development. It is assumed that registrants are already familiar with the fundamental optimization techniques.
Domain-Specific Languages
Computer graphics has introduced several domain-specific languages (DSLs) that enable high performance and parallelism for narrow problem domains - RenderMan, Cg, GLSL, and recently OpenRL and OptiX. We think that similar approaches can benefit other areas of GPU computing - visualization, animation, physics simulation, or scientific data analysis. In this talk, we present Shadie, a domain-specific shading language for rapid development of complex custom volume visualizations in radiation oncology. The shaders are written in a high-level Python-like language and translated to CUDA for efficiency. We will explain how you can develop your own DSLs using source-to-source translation and a suitable backend library.
In addition to presenting submissions at the GPU Technology Conference, selected speakers may have the opportunity to showcase their work at future upcoming NVIDIA supported events.
To be notified of Call for Submissions details and deadlines, sign up for our email announcements.