Towards GPU Utilization Prediction for Cloud Deep Learning

Authors: 

Gingfung Yeung, Damian Borowiec, Adrian Friday, Richard Harper, and Peter Garraghan, Lancaster University

Abstract: 

Understanding the GPU utilization of Deep Learning (DL) workloads is important for enhancing resource-efficiency and cost-benefit decision making for DL frameworks in the cloud. Current approaches to determine DL workload GPU utilization rely on online profiling within isolated GPU devices, and must be performed for every unique DL workload submission resulting in resource under-utilization and reduced service availability. In this paper, we propose a prediction engine to proactively determine the GPU utilization of heterogeneous DL workloads without the need for in-depth or isolated online profiling. We demonstrate that it is possible to predict DL workload GPU utilization via extracting information from its model computation graph. Our experiments show that the prediction engine achieves an RMSLE of 0.154, and can be exploited by DL schedulers to achieve up to 61.5% improvement to GPU cluster utilization.

Open Access Media

USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are posted after the event are also free and open to everyone. Support USENIX and our commitment to Open Access.

BibTeX
@inproceedings {254126,
author = {Gingfung Yeung and Damian Borowiec and Adrian Friday and Richard Harper and Peter Garraghan},
title = {Towards {GPU} Utilization Prediction for Cloud Deep Learning},
booktitle = {12th {USENIX} Workshop on Hot Topics in Cloud Computing (HotCloud 20)},
year = {2020},
url = {https://www.usenix.org/conference/hotcloud20/presentation/yeung},
publisher = {{USENIX} Association},
month = jul,
}

Presentation Video