Understanding Training Efficiency of Deep Learning Recommendation Models at Scale
Authors:
- 1 February 2021
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 802-814
Abstract
The use of GPUs has proliferated for machine learning workflows and is now considered mainstream for many deep learning models. Meanwhile, when training state-of-the-art personal recommendation models, which consume the highest number of compute cycles at our large-scale datacenters, the use of GPUs came with various challenges due to having both compute-intensive and memory-intensive components. GPU performance and efficiency of these recommendation models are largely affected by model architecture configurations such as dense and sparse features, MLP dimensions. Furthermore, these models often contain large embedding tables that do not fit into limited GPU memory. The goal of this paper is to explain the intricacies of using GPUs for training recommendation models, factors affecting hardware efficiency at scale, and learnings from a new scale-up GPU server design, Zion.Keywords
All Related Versions
This publication has 19 references indexed in Scilit:
- Focal Loss for Dense Object Detection2017
- Mask R-CNN2017
- Wide & Deep Learning for Recommender Systems2016
- Deep Neural Networks for YouTube Recommendations2016
- Cross-architecture performance prediction (XAPP) using CPU code to predict GPU performance2015
- A Multi-View Deep Learning Approach for Cross Domain User Modeling in Recommendation Systems2015
- Predicting GPU Performance from CPU Runs Using Machine Learning2014
- The tail at scaleCommunications of the ACM, 2013
- Data warehousing and analytics infrastructure at facebook2010
- RooflineCommunications of the ACM, 2009