2017 IEEE High Performance Extreme Computing Conference (HPEC ‘17) Twenty-first Annual HPEC Conference 12 - 14 September 2017 Westin Hotel, Waltham, MA USA
Site created and maintained by Ballos Associates
Tutorial: GPU Computing – Introduction to CUDA & Deep Learning
Instructor:  Dr. Larry Brown, Solutions Architect, NVIDIA (9am-12:30pm) 9:00 – 10:45 Community Detection Using nvGRAPH in CUDA 9.0 With the CUDA 9.0 release, nvGRAPH includes maximum modularity clustering, a common technique for community detection.  We will walk through some sample code to connect a given graph to nvGRAPH and calculate maximum modularity partitions of that graph.  Graph contraction can also be used with a Jaccard metric to find strongly connected neighborhoods very quickly. We will demonstrate how to set up nvGRAPH’s graph contraction features, and run graph contraction to produce both clusters and a new, smaller graph.  Next, we will show the new BFS code, and combine graph contraction with BFS to run connected components or SSSP faster than possible on the initial graph.   On completion of this tutorial you will have a working knowledge of how to apply nvGRAPH to large graph analytics problems.   11:00 – 12:30 Image Classification with DIGITS Deep learning is giving machines near human levels of visual recognition capabilities and disrupting many applications by replacing hand-coded software with predictive models learned directly from data. This lab introduces the machine learning workflow and provides hands-on experience with using deep neural networks (DNN) to solve a real-world image classification problem. You will walk through the process of data preparation, model definition, model training and troubleshooting, validation testing and strategies for improving model performance. You will also see the benefits of GPU acceleration in the model training process. On completion of this lab you will have the knowledge to use NVIDIA DIGITS to train a DNN on your own image classification dataset.