This SCGC-FIRST project develops new materials to enable energy-efficient artificial intelligence through reservoir computing. As machine learning systems consume increasing amounts of energy, alternative computing architectures are urgently needed. Reservoir computing offers comparable processing power to conventional approaches but with far fewer parameters, significantly reducing energy demands.
At the heart of this approach is a “physical reservoir” — a dynamic material system in which many interacting nodes continuously evolve. The project focuses on advanced electrolyte materials can be used to improve the accuracy and performance of reservoir computing devices.
The research will involve the organic synthesis of novel electrolytes, followed by detailed analysis of their physical properties. By identifying materials with enhanced conductivity and tailored physical properties, the project aims to unlock smaller, more accurate, and more energy-efficient reservoir computing devices, supporting sustainable innovation in AI technologies.
