Fred Streitz of Lawrence Livermore National Laboratory (LLNL) leads efforts to develop supercomputer applications that address forefront scientific problems by pushing the limits of leadership-class computing. He is also director of LLNL’s High-Performance Computing Innovation Center (HPCIC), where he develops and strengthens collaborations with academic and industrial partners. Further, he is a Fellow of the American Physical Society and a two-time winner of the Gordon Bell Prize for significant achievement in supercomputing.
At the SC18 supercomputing conference in Dallas, he gave a talk in the US Department of Energy booth on the topic “Machine Learning and Predictive Simulation: HPC and the US Cancer Moonshot on Sierra.” As a guest on the Exascale Computing Project podcast, he provides an overview and some insights from his booth talk.
Delving into a particular type of cancer [1:39]
RAS protein mutation: a hallmark of some of the most vicious cancers [2:01]
Using the Sierra supercomputer in partnership with the National Cancer Institute to understand RAS protein behavior in a lipid [3:39]
The purpose of this project [4:12]
Initial exploration of some of computing’s forefront workflow challenges [5:06]
Experimentation and simulation with machine learning in the middle [6:53]
Using machine learning to optimize decision-making [9:25]
The project deliverable [12:06]