Authors: Florentia Afentaki, Michael Hefenbrock, Georgios Zervakis, Mehdi B. Tahoori

Published on: February 05, 2024

Impact Score: 8.52

Arxiv code: Arxiv:2402.0293

Summary

• What is new: Integration of hardware approximation into the MLP training process for Printed Electronics.
• Why this is important: Difficulty in implementing complex circuits in Printed Electronics due to larger feature sizes.
• What the research proposes: A genetic-based, approximate, hardware-aware training approach for MLPs in PE.
• Results: Over 5x area and power reduction with only a 5% accuracy loss, outperforming current approximate and stochastic MLPs.

Technical Details

Technological frameworks used: nan

Models used: Multilayer Perceptrons (MLPs)

Data used: nan

Potential Impact

Companies in flexible electronics and wearable technology markets.

Want to implement this idea in a business?

We have generated a startup concept here: FlexAI.