Treffer: Energy - efficient AI models for embedded systems: Innovations in low - power computing.

Title:
Energy - efficient AI models for embedded systems: Innovations in low - power computing.
Authors:
Vij, Priya1 (AUTHOR) ku.priyavij@kalingauniversity.ac.in, Prashant, Patil Manisha1 (AUTHOR) patil.manisha@kalingauniversity.ac.in
Source:
AIP Conference Proceedings. 2026, Vol. 3345 Issue 1, p1-6. 6p.
Database:
Academic Search Index

Weitere Informationen

As digital systems get a more essential position in today's modern technology, demand for energy economical AI models is increasing, thus calling for developing low-power computing. These approaches have significant problems of high power and low computational efficiency that affect their performance where the resources are limited. For these reasons, we propose the Low-Power Computing using Artificial Intelligence (LPC-AI) approach that involves leveraging advanced optimization techniques and also energy-aware algorithms. The LPC-AI model improves power efficiency while maintaining high computational performance, which makes it suitable for deployment in embedded systems with tight energy budgets. Our findings show that the LPC-AI model reduces power consumption significantly compared to traditional methods by up to 40% while improving processing speed by 25%. This innovation represents a colossal stride toward making AI a reality in applications with very little or no electrical consumption. [ABSTRACT FROM AUTHOR]