Skip to content
CaMLSysUniversity of Cambridge
Home People Publications Projects Contact

← All publications

Best of Both Worlds: AutoML Codesign of a CNN and its Hardware Accelerator

  • Mohamed S. AbdelfattahSamsung AI Center, Cambridge
  • Lukasz DudziakSamsung AI Center, Cambridge
  • Thomas ChauSamsung AI Center, Cambridge
  • Royson LeeSamsung AI Center, Cambridge
  • Hyeji KimSamsung AI Center, Cambridge
  • Nicholas D. LaneSamsung AI Center, CambridgeUniversity of Oxford

ACM/IEEE Design Automation Conference (DAC), 2020

Abstract

Neural architecture search (NAS) has been very successful at outperforming human-designed convolutional neural networks (CNN) in accuracy, and when hardware information is present, latency as well. However, NAS-designed CNNs typically have a complicated topology, therefore, it may be difficult to design a custom hardware (HW) accelerator for such CNNs. We automate HW-CNN codesign using NAS by including parameters from both the CNN model and the HW accelerator, and we jointly search for the best model-accelerator pair that boosts accuracy and efficiency. We call this Codesign-NAS. In this paper we focus on defining the Codesign-NAS multiobjective optimization problem, demonstrating its effectiveness, and exploring different ways of navigating the codesign search space. For CIFAR-10 image classification, we enumerate close to 4 billion model-accelerator pairs, and find the Pareto frontier within that large search space. This allows us to evaluate three different reinforcement-learning-based search strategies. Finally, compared to ResNet on its most optimal HW accelerator from within our HW design space, we improve on CIFAR-100 classification accuracy by 1.3% while simultaneously increasing performance/area by 41% in just~1000 GPU-hours of running Codesign-NAS.

Materials

  • PDF
BibTeX
@inproceedings{abdelfattah2020best,
  title = {{Best of Both Worlds: AutoML Codesign of a CNN and its Hardware Accelerator}},
  author = {Abdelfattah, Mohamed S. and Dudziak, Lukasz and Chau, Thomas and Lee, Royson and Kim, Hyeji and Lane, Nicholas D.},
  year = {2020},
  booktitle = {ACM/IEEE Design Automation Conference (DAC)},
  pages = {1--6},
  publisher = {IEEE},
  doi = {10.1109/dac18072.2020.9218596},
  url = {https://doi.org/10.1109/dac18072.2020.9218596}
}
Cambridge Machine Learning Systems Group
Department of Computer Science and Technology
University of Cambridge
ContactGitHub