Collaborative Research: ENG-SEMICON: RUI: Carbohydrate Memristor Empowered Environmentally Sustainable Processor-in-Memory

Carbohydrate Memristor Empowered Environmentally Sustainable Processor-in-Memory Nontechnical description:

Artificial intelligence (AI) systems have profound influence on societal wellbeing of humans and fueling significant economic growth. However, operation of conventional computing architecture in AI systems as well as manufacturing and disposal of conventional computing devices lead to significant energy consumption, depletion of nonrenewable natural resources, and ecological deterioration.

Therefore, a serious concern of environmental sustainability to such increasingly pervasive computing systems has been raised, and improvements in system performance, energy efficiency, and ecological friendliness require new devices and systems. In this project, a new environmentally-sustainable processor-in-memory system is proposed to benefit the entire computing community including mobile and wearable computing, cloud computing and data center, electronic sensing and controlling, communication and networking.

This project will also contribute to the development of high-quality workforce skilled in design, fabrication, testing, and modeling of memory devices and processor-in-memory computing systems for the growing needs in the US. The students and postdoctoral scholar participating in this project will receive unique training in engineering problem solving and technology development, and their research and educational experience will be enhanced by complementary expertise and close collaboration between the two research groups.

Technical description:

Processor-in-memory systems implemented with memristors have great potential to perform complex AI computations faster and on a smaller footprint. The goal of this project is to address the environmental sustainability challenge in computing by developing a novel brain-inspired processor-in-memory system empowered by memristors made from carbohydrate materials for energy-efficient operation, renewable material resource, sustainable device manufacturing, and ecologically-friendly disposal.

The carbohydrate materials will be naturally extracted from plants, vegetable, and fruits with low cost and waste generation. Innovative fabrication techniques for carbohydrate-based memristor and processor-in-memory system will be developed with reduced water use and chemical waste, greenhouse gas emission, and manufacturing related energy consumption.

The processor-in-memory system will be constructed by implementing carbohydrate memristors and reconfigurable peripheral circuits to achieve sustainable computing, enable performance improvement, and ensure high operation efficiency, longevity, and reliability.

This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.