Repository logo

Infoscience

  • English
  • French
Log In
Logo EPFL, École polytechnique fédérale de Lausanne

Infoscience

  • English
  • French
Log In
  1. Home
  2. Academic and Research Output
  3. Conferences, Workshops, Symposiums, and Seminars
  4. Backpropagation-free Training of Analog AI Accelerators
 
Loading...
Thumbnail Image
conference paper

Backpropagation-free Training of Analog AI Accelerators

Momeni, A.  
•
Rahmani, Babak
•
Mallejac, Matthieu  
Show more
2024
2024 18th International Congress on Artificial Materials for Novel Wave Phenomena, Metamaterials 2024
18 International Congress on Artificial Materials for Novel Wave Phenomena

Deep learning has achieved remarkable success in diverse fields in recent years. However, this growth presents significant challenges, particularly in terms of energy consumption during both training and inference phases. While there have been efforts to improve energy efficiency during the inference phase, efficient training of deep learning models remains a largely unaddressed challenge. The training method for digital deep learning models typically relies on backpropagation, a process that is difficult to implement physically due to its reliance on precise knowledge of forward-pass computations in neural networks. To overcome this issue, We present a physics-compatible deep neural network architecture, augmented by a biologically-inspired learning algorithm referred to as physical local learning (PhyLL). This framework allows for the direct training of deep physical neural networks, comprising layers of physical nonlinear systems. Notably, our approach dispenses with the need for detailed knowledge of the specific properties of these nonlinear physical layers. Our approach outperforms state-of-the-art hardware-aware training methods by enhancing training speed, reducing digital computations and power consumption in physical systems, particularly in optics.

  • Details
  • Metrics
Type
conference paper
DOI
10.1109/Metamaterials62190.2024.10703263
Scopus ID

2-s2.0-85207820280

Author(s)
Momeni, A.  
•
Rahmani, Babak
•
Mallejac, Matthieu  
•
Hougne, Philipp Del
•
Fleury, Romain  
Date Issued

2024

Publisher

Institute of Electrical and Electronics Engineers Inc.

Published in
2024 18th International Congress on Artificial Materials for Novel Wave Phenomena, Metamaterials 2024
ISBN of the book

9798350373493

Peer reviewed

REVIEWED

Written at

EPFL

EPFL units
LWE  
Event nameEvent acronymEvent placeEvent date
18 International Congress on Artificial Materials for Novel Wave Phenomena

Chania, Greece

2024-09-09 - 2024-09-14

Available on Infoscience
January 26, 2025
Use this identifier to reference this record
https://infoscience.epfl.ch/handle/20.500.14299/244888
Logo EPFL, École polytechnique fédérale de Lausanne
  • Contact
  • infoscience@epfl.ch

  • Follow us on Facebook
  • Follow us on Instagram
  • Follow us on LinkedIn
  • Follow us on X
  • Follow us on Youtube
AccessibilityLegal noticePrivacy policyCookie settingsEnd User AgreementGet helpFeedback

Infoscience is a service managed and provided by the Library and IT Services of EPFL. © EPFL, tous droits réservés