Addressing the Memory Bottleneck in AI Model Training

David Ojika; Bhavesh Patel; G. Anthony Reina; Trent Boyer; Chad Martin; Prashant Shah

Addressing the Memory Bottleneck in AI Model Training

Machine Learning 2020-03-20 v1 Computer Vision and Pattern Recognition Image and Video Processing

Authors: David Ojika , Bhavesh Patel , G. Anthony Reina , Trent Boyer , Chad Martin , Prashant Shah

Abstract

Using medical imaging as case-study, we demonstrate how Intel-optimized TensorFlow on an x86-based server equipped with 2nd Generation Intel Xeon Scalable Processors with large system memory allows for the training of memory-intensive AI/deep-learning models in a scale-up server configuration. We believe our work represents the first training of a deep neural network having large memory footprint (~ 1 TB) on a single-node server. We recommend this configuration to scientists and researchers who wish to develop large, state-of-the-art AI models but are currently limited by memory.

Keywords

deep neural network acceleration large language model inference deep learning

Cite

@article{arxiv.2003.08732,
  title  = {Addressing the Memory Bottleneck in AI Model Training},
  author = {David Ojika and Bhavesh Patel and G. Anthony Reina and Trent Boyer and Chad Martin and Prashant Shah},
  journal= {arXiv preprint arXiv:2003.08732},
  year   = {2020}
}

Comments

Presented at Workshop on MLOps Systems at MLSys 2020 Conference, Austin TX

Addressing the Memory Bottleneck in AI Model Training

Abstract

Keywords

Cite

Comments

Related papers