English

Enabling High Data Throughput Reinforcement Learning on GPUs: A Domain Agnostic Framework for Data-Driven Scientific Research

Machine Learning 2024-08-05 v1 Artificial Intelligence

Abstract

We introduce WarpSci, a domain agnostic framework designed to overcome crucial system bottlenecks encountered in the application of reinforcement learning to intricate environments with vast datasets featuring high-dimensional observation or action spaces. Notably, our framework eliminates the need for data transfer between the CPU and GPU, enabling the concurrent execution of thousands of simulations on a single or multiple GPUs. This high data throughput architecture proves particularly advantageous for data-driven scientific research, where intricate environment models are commonly essential.

Keywords

Cite

@article{arxiv.2408.00930,
  title  = {Enabling High Data Throughput Reinforcement Learning on GPUs: A Domain Agnostic Framework for Data-Driven Scientific Research},
  author = {Tian Lan and Huan Wang and Caiming Xiong and Silvio Savarese},
  journal= {arXiv preprint arXiv:2408.00930},
  year   = {2024}
}
R2 v1 2026-06-28T18:01:37.708Z