Researchers at the University of Cambridge have developed a new nanoelectronic device that could significantly reduce the energy demands of artificial intelligence (AI) hardware by mimicking the way the human brain processes information.
The breakthrough centres on a novel form of hafnium oxide that functions as a highly stable, low-energy memristor — a component designed to replicate how neurons connect and communicate. The findings, published in Science Advances, point to a potential step-change in how AI systems are built and powered.
Current AI architectures rely on conventional semiconductor chips that continuously transfer data between memory and processing units. This back-and-forth movement is highly energy intensive, a growing concern as global AI adoption accelerates across sectors.
The Cambridge team’s approach draws on neuromorphic computing, a brain-inspired model that integrates memory and processing in the same location. This could cut energy consumption by up to 70%, while also enabling systems to learn and adapt more efficiently.
“Energy consumption is one of the key challenges in current AI hardware,” said Dr Babak Bakhit, lead author from Cambridge’s Department of Materials Science and Metallurgy. He added that next-generation devices must combine ultra-low power operation with stability and the ability to switch between multiple states.
Traditional memristors typically rely on conductive filaments within metal oxides to function. However, these filaments can behave unpredictably and often require high operating voltages, limiting their scalability.
In contrast, the Cambridge researchers engineered a hafnium-based thin film incorporating strontium and titanium, produced using a two-step growth process. This created nanoscale ‘p-n junctions’ at the interface between layers, enabling the device to switch states by modulating an energy barrier rather than forming filaments.
According to Bakhit, this interface-driven mechanism delivers far greater consistency. “Filamentary devices suffer from random behaviour,” he said. “But because our devices switch at the interface, they show outstanding uniformity from cycle to cycle and from device to device.”
The devices demonstrated switching currents around a million times lower than some conventional alternatives, while also supporting hundreds of stable conductance levels — a critical feature for analogue, in-memory computing.
Laboratory tests showed the memristors could endure tens of thousands of switching cycles and retain programmed states for up to a day. They also replicated key biological learning behaviours, including spike-timing dependent plasticity, where neural connections strengthen or weaken based on signal timing.
Despite the promise, challenges remain. The fabrication process currently requires temperatures of around 700°C, exceeding typical semiconductor manufacturing limits. The team is now working to lower this threshold to enable integration into commercial chip production.
“If we can reduce the temperature and put these devices onto a chip, it would be a major step forward,” Bakhit said.
A patent application has been filed through Cambridge Enterprise, with the research supported by organisations including UK Research and Innovation and the Royal Society.
Image credit: University of Cambridge
Looking for your next technology, science or engineering role? Maybe you’re just starting out in your career? Polytec is here to help you find the perfect position for you. Browse our current vacancies now >>> www.polytec.co.uk/jobs
Researchers at the University of Cambridge have developed a new nanoelectronic device that could significantly reduce the energy demands of artificial intelligence (AI) hardware by mimicking the way the human brain processes information.
The University of Cambridge is set to host the UK’s most powerful quantum computer as part of a major new partnership with US-based quantum technology firm IonQ, marking the University’s largest-ever corporate research collaboration.
The University of Cambridge has been ranked the UK’s top employer in the Financial Times Best Employers 2026 survey, a nationwide assessment based on anonymous employee feedback.
Researchers from the University of Cambridge are among a global team developing a new generation of artificial intelligence models trained not on text or images, but on real-world scientific data, in a move they say could accelerate discovery across disciplines.
A University of Cambridge spinout developing photon-multiplier technology to boost the efficiency of silicon solar panels, has secured a £1.56m pre-Series A investment to accelerate commercial readiness of its breakthrough energy technology.
Developers behind the proposed Cambridge 25 employment park near Junction 25 of the A14 say updated plans could deliver significant economic and community benefits for Greater Cambridge, as a fresh public consultation opens ahead of a spring planning submission.
A vast new employment park could transform Huntingdonshire’s economy, with developers claiming it has the potential to create up to 15,000 jobs.
Cambridge Dictionary has added more than 6,000 new words, phrases and meanings in the past year, with entries ranging from internet slang to political portmanteaux.
The British public is divided over how artificial intelligence should be used in schools, with strong support for its role in helping teachers with administrative work but widespread opposition to its use for marking.
New figures from CV-Library reveal a striking divergence in the UK’s AI job market, with robotics-related positions surging ahead in pay growth while many traditional AI roles experience steep declines.
Oxford and Cambridge universities have each received a £6.25m donation from an anonymous benefactor to expand outreach and support for disadvantaged students in science, technology, engineering and mathematics (STEM).
Meta has launched a new £12m audio research facility in Cambridge as part of its global push to develop advanced AI-powered smart glasses.