Attending this event?
Back To Schedule
Monday, October 25 • 11:40am - 12:20pm
Scalable Deep Learning for Large Scale Scientific Machine Learning: Challenges and Opportunities

Sign up or log in to save this to your schedule, view media, leave feedback and see who's attending!

Log in to leave feedback.
Sponsored by SambaNova Systems

Scientific Machine Learning at HPC scale presents many challenges related to massive data sets, applications with large sample sizes, and large neural network models. Examples of these challenges are showcased in application areas ranging from small molecule drug design, to cosmology, and sequence-based transformer language models. In this talk we discuss the opportunities that these challenge create and present our on-going work in developing and composing multiple methods of parallel training within the LBANN scalable deep learning toolkit for these SciML applications. Furthermore, we showcase how these methods have enabled the use of leadership-class HPC systems, such as Sierra, to accelerate the training of neural network architectures.

avatar for Brian Van Essen

Brian Van Essen

Informatics Group leader and Computer Scientist, Center for Applied Scientific Computing, Lawrence Livermore National Laboratory (LLNL)
Brian is the Informatics Group leader and a Computer Scientist in the Center for Applied Scientific Computing at Lawrence Livermore National Laboratory (LLNL). He is actively pursuing research in large-scale deep learning for scientific domains and training deep neural networks using... Read More →

Monday October 25, 2021 11:40am - 12:20pm CDT

Attendees (8)