SC20 Is Everywhere We Are

SC20 Virtual Platform
PREEMPT: Scalable Epidemic Interventions Using Submodular Optimization on Multi-GPU Systems
Event Type
Accelerators, FPGA, and GPUs
Graph Algorithms
Scalable Computing
Registration Categories
TimeWednesday, 18 November 20201pm - 1:30pm EDT
LocationTrack 3
DescriptionPreventing and slowing the spread of epidemics are achieved through techniques such as vaccination and social distancing. Given practical limitations on the number of vaccines and the cost of administration, optimization becomes a necessity. Previous approaches using mathematical programming methods have been shown to be effective but are limited by computational costs. In this work, we make several contributions: First, we present a new approach to intervention via maximizing the influence of vaccinated nodes on the network. We call this method PREEMPT. Next, we present a new parallel algorithm for PREEMPT, along with a CPU-GPU implementation. Our results demonstrate that PREEMPT is able to achieve significant speedups on Summit while improving the quality of solutions. This work represents a first-of-its-kind effort in parallelizing greedy hill climbing and applying it toward devising effective interventions for epidemics.
Back To Top Button