HPC and AI/ML: A Synergistic Relationship

Explore the synergistic relationship between High Performance Computing (HPC) and Artificial Intelligence/Machine Learning (AI/ML) in this 59-minute conference talk by Abhinav Bhatele, presented at SPCL_Bcast. Discover how the increasing memory capacity and computational power of modern architectures have fueled the training of extremely large deep neural networks. Learn about AxoNN, a parallel deep learning framework that optimizes work scheduling and communication through asynchrony and message-driven execution. Examine how neural network properties can be leveraged for systems-focused optimizations. Delve into the application of AI/ML in analyzing HPC system monitoring data, including examples of data-driven ML models for performance modeling, forecasting, and code generation. Gain insights into the convergence of HPC and AI/ML fields and their mutual benefits through this comprehensive presentation, which includes an introduction, the main talk, announcements, and a Q&A session.

Intro
Talk
Announcements
Q&A