Communications of the ACM
Project Aims to Optimize Workflows to Amplify AI for Science
Poseidon will develop deep-reinforcement learning methods that can self-learn corrective behaviors and optimize workflow performance.
Scientists who come to the U.S. Department of Energy's national laboratories to solve big problems are turning increasingly to artificial intelligence and machine learning to help answer scientific questions. As AI and ML continue to scale and advance, so does the complexity of running them on supercomputers and distributed computing networks.
Scientists at the DOE's Argonne National Laboratory are tackling this challenge by modeling, simulating, predicting and optimizing the performance of workflows. The Platform for Explainable Distributed Infrastructure (Poseidon) project, funded by the DOE, is turning to AI and ML to improve the performance of these workflows.
Poseidon aims to advance the knowledge of how simulation and ML methodologies can be harnessed and amplified to improve DOE's computational and data science applications.
"By optimizing science workflows that run on distributed computing and data infrastructure, we will be able to accelerate scientific discovery," says Prasanna Balaprakash, a computer science leader at Argonne. Balaprakash's team and Argonne's Poseidon partners will model science workflows, predict their performances, automatically identify performance anomalies, and optimize entire workflows to ensure they run as fast and efficiently as possible.
From Argonne National Laboratory
View Full Article
No entries found