This new brain-like chip could slash AI energy use by 70%

The research team, led by the University of Cambridge, developed a modified version of hafnium oxide that functions as a highly stable, low-energy 'memristor' -- a component designed to replicate how neurons connect and communicate in the brain. Their findings were published in the journal Science Advances.

Why Current AI Systems Use So Much Energy

Modern AI relies on traditional computer chips that constantly move data between memory and processing units. This back-and-forth transfer requires large amounts of electricity, and demand continues to rise as AI becomes more widely used across industries.

Neuromorphic computing offers a different approach. Instead of separating memory and processing, it combines both in one place, similar to how the brain works. This method could cut energy use by as much as 70% while also allowing systems to learn and adapt more naturally.

"Energy consumption is one of the key challenges in current AI hardware," said lead author Dr. Babak Bakhit, from Cambridge's Department of Materials Science and Metallurgy. "To address that, you need devices with extremely low currents, excellent stability, outstanding uniformity across switching cycles and devices, and the ability to switch between many distinct states."

A New Approach to Memristor Design

Most existing memristors operate by forming tiny conductive filaments inside metal oxide materials. These filaments tend to behave unpredictably and often require high voltages, which limits their practicality for large-scale computing.

The Cambridge researchers took a different route. They engineered a hafnium-based thin film that switches states through a more controlled mechanism. By adding strontium and titanium and using a two-step growth process, they created small electronic gates, known as 'p-n junctions', at the interfaces between layers.

Instead of relying on filaments forming and breaking, the device changes its resistance by adjusting the energy barrier at these interfaces. This allows for smoother and more reliable switching.

Bakhit, who is also affiliated with Cambridge's Department of Engineeirng, explained that this design solves a major issue in memristor development. "Filamentary devices suffer from random behavior," he said. "But because our devices switch at the interface, they show outstanding uniformity from cycle to cycle and from device to device."

Ultra-Low Power and Brain-Like Learning

Tests showed that the new devices operate at switching currents roughly a million times lower than some conventional oxide-based memristors. They can also achieve hundreds of stable conductance levels, which is essential for analogue 'in-memory' computing.

In laboratory experiments, the devices remained stable through tens of thousands of switching cycles and retained their programmed states for about a day. They also demonstrated key biological learning behaviors, including spike-timing dependent plasticity: the process that allows neurons to strengthen or weaken their connections based on timing.

"These are the properties you need if you want hardware that can learn and adapt, rather than just store bits," said Bakhit.

Remaining Challenges and Future Potential

Despite the promising results, there are still obstacles to overcome. The current manufacturing process requires temperatures of around 700°C -- higher than what standard semiconductor fabrication typically allows.

"This is currently the main challenge in our device fabrication process," said Bakhit. "But we're now working on ways to bring the temperature down to make it more compatible with standard industry processes."

If this issue can be resolved, the technology could be integrated into practical chip-scale systems. "If we can reduce the temperature and put these devices onto a chip, it would be a major step forward," he said.

Years of Trial and Error Behind the Breakthrough

The advance came after several years of experimentation and many setbacks. Bakhit said progress finally accelerated late last year when he modified the fabrication process, adding oxygen only after forming the first layer.

"I spent almost three years on this," he said. "There were a huge number of failures. But at the end of November, we saw the first really good results. It's still early days of course, but if we can solve the temperature issue, this technology could be game-changing because the energy consumption is so much lower and at the same time, the device performance is highly promising."

The work was supported in part by the Swedish Research Council (VR), the Royal Academy of Engineering, the Royal Society, and UK Research and Innovation (UKRI). A patent application has been filed by Cambridge Enterprise, the University's innovation arm.