EU-40 Materials prize

Resistance Switching in Oxide-Based Materials for Future
Energy-Efficient Memory and Computing

Adnan Mehonic
Co-founder and CTO, Intrinsic Semiconductor Technologies
University College London
Department of Electronic & Electrical Engineering
London, United Kingdom

The increasing demand for computing power has led to unsustainable energy consumption. By 2035, the energy required for computing systems is projected to exceed realistic supply, constraining the development of transformative applications driven by complex algorithms. Memory is the cornerstone of computing efficiency: storage access and data transfer account for most of the energy consumed in advanced systems. Beyond the largest, most complex, energy-hungry systems, there is a pressing need to embed intelligence in devices that must operate on extremely low power budgets - an advance that would transform numerous applications.

This talk traces a 15-year effort to develop a memory technology that meets this challenge. What began as a materials-engineering study of resistance switching in amorphous oxides has matured into a resistive RAM (ReRAM) platform that addresses the requirements for non-volatile storage and novel efficient computing hardware. I will outline contributions to understanding the resistance switching physics, tuning performance through materials engineering and device design, and designing reliable, manufacturable technology. The presentation concludes with a roadmap that defines ReRAM’s role in future memory-centric and novel computing systems, together with lessons learned while translating the technology from academia to industry, including the pathway pursued through the spin-out whose goal is to bridge the gap between innovation in materials engineering and widespread application of novel technology that promises to be the game-changer for future efficient memory and computing systems.

Biography:

Dr. Adnan Mehonic is an Associate Professor of Nanoelectronics and Royal Academy of Engineering Senior Research Fellow at University College London and the Co-founder and Chief Technology Officer of Intrinsic Semiconductor Technologies. His research spans energy-efficient nanoelectronic devices and systems, encompassing fundamental materials science, novel device design, and memory and efficient computing hardware design.

He pioneered the development of oxide-based memristors, providing key insights into switching mechanisms, device-, algorithm- and system-level designs. Adnan has published more than 100 journal articles and conference papers and is named on 15 international patents. He has been recognised by inclusion in MIT Technology Review’s “35 Innovators Under 35,” Wiley’s “Rising Star,” and UCL’s “One to Watch.”

Alongside his academic career, he has successfully translated academic research into industrial impact, securing over £15 million in venture funding for Intrinsic Semiconductor Technologies, building a successful start-up consisting of a team of 15 highly skilled engineers and scientists, and developing mature industry prototypes through multiple collaborations with international industrial centres. He is the founding Editor-in-Chief of APL Machine Learning, where he fosters collaboration between applied physics and artificial intelligence communities, and a co-director of NeuMat, the Neuromorphic Network in the UK, dedicated to supporting and uniting researchers, industry professionals, and innovators working on neuromorphic technologies for advanced, novel, and energy-efficient AI hardware.