A GPU cluster for empowering research on distributed optimization and learning in massive-scale systems

Abstract

Non-convex, discrete, and stochastic optimization models are key to solving decision-making problems in a variety of contexts, and also form the backbone for modern machine learning techniques. However, traditional computing architectures are not sufficiently scalable to solve truly large-scale models arising in many DoD applications. The research team, consisting of a PI and six (6) Co-PIs currently working on 13 projects funded by several DoD sources (AFOSR, ONR and DARPA), seek to use modern computing platforms based on Graphical Processing Units (GPUs). The current projects by the PIs involve the deployment of advanced optimization and machine learning techniques in theaters which evolve rapidly, with adversarial and uncertain environments. With the acquisition of the proposed equipment, the PIs will transition their methods from current CPU-based architectures to massively parallel architectures available with modern GPU platforms.

Document Details

Document Type
DoD Grant Award
Publication Date
Mar 06, 2024
Source ID
FA95502310551

Entities

People

  • Andres Gomez Escobar

Organizations

  • Air Force Office of Scientific Research
  • United States Air Force
  • University of Southern California

Tags

Fields of Study

  • Computer science

Readers

  • Distributed Systems and Data Platform Development
  • Parallel and Distributed Computing.

Technology Areas

  • AI & ML
  • AI & ML - Machine Learning Algorithms
  • AI & ML - Neural Networks