EUVIMEDEUVIMED entdecken
Open Access · CC BY 4.0

SpikeExplorer: Hardware-Oriented Design Space Exploration for Spiking Neural Networks on FPGA

Dario Padovano, Alessio Carpegna, Alessandro Savino, Stefano Di Carlo

Electronics · 01.05.2024 · 13 (9) · 1744

DOI: 10.3390/electronics13091744 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Electronics, online erschienen am 01.05.2024. Verfasst von Dario Padovano und 3 weiteren Autor:innen. Die Arbeit wurde bisher 9-mal zitiert (Crossref) und nennt 46 Literaturangaben. Der Volltext ist frei zugänglich (Open Access, CC BY 4.0). Gefördert von: European Union, MUR.

AbstractBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

One of today’s main concerns is to bring artificial intelligence capabilities to embedded systems for edge applications. The hardware resources and power consumption required by state-of-the-art models are incompatible with the constrained environments observed in edge systems, such as IoT nodes and wearable devices. Spiking Neural Networks (SNNs) can represent a solution in this sense: inspired by neuroscience, they reach unparalleled power and resource efficiency when run on dedicated hardware accelerators. However, when designing such accelerators, the amount of choices that can be taken is huge. This paper presents SpikExplorer, a modular and flexible Python tool for hardware-oriented Automatic Design Space Exploration to automate the configuration of FPGA accelerators for SNNs. SpikExplorer enables hardware-centric multiobjective optimization, supporting target factors such as accuracy, area, latency, power, and various combinations during the exploration process. The tool searches the optimal network architecture, neuron model, and internal and training parameters leveraging Bayesian optimization, trying to reach the desired constraints imposed by the user. It allows for a straightforward network configuration, providing the full set of explored points for the user to pick the trade-off that best fits their needs. The potential of SpikExplorer is showcased using three benchmark datasets. It reaches 95.8% accuracy on the MNIST dataset, with a power consumption of 180 mW/image and a latency of 0.12 ms/image, making it a powerful tool for automatically optimizing SNNs.

Förderung

  • European Union
  • MUR

Zitieren

APA

Padovano, D., Carpegna, A., Savino, A., & Carlo, S. D. (2024). SpikeExplorer: Hardware-Oriented Design Space Exploration for Spiking Neural Networks on FPGA. Electronics, 13(9), 1744. https://doi.org/10.3390/electronics13091744

Vancouver

Padovano D, Carpegna A, Savino A, Carlo SD. SpikeExplorer: Hardware-Oriented Design Space Exploration for Spiking Neural Networks on FPGA. Electronics. 2024;13(9):1744. doi:10.3390/electronics13091744

BibTeX
@article{Padovano2024,
  author = {Padovano, Dario and Carpegna, Alessio and Savino, Alessandro and Carlo, Stefano Di},
  title = {SpikeExplorer: Hardware-Oriented Design Space Exploration for Spiking Neural Networks on FPGA},
  journal = {Electronics},
  year = {2024},
  volume = {13},
  number = {9},
  pages = {1744},
  doi = {10.3390/electronics13091744}
}
Kostenlos starten →