(Image: Professor Andreas Waag (TU Braunschweig) and Professor Christian Werner (Ostfalia) at the first demonstrator of the neuromorphic computer design)

They want to bring light to the world of microelectronics: The state of Lower Saxony and the Volkswagen Foundation are providing around €15 million in funding to the Lower Saxony microelectronics consortium ‘Nitride Technology Centre’ (NTC), which is led by TU Braunschweig. Under the project name ‘Bringing Light to Microelectronics’ (BRIGHT), the consortium is developing, among other things, a unique neuromorphic computer based on LEDs. In the future, this is expected to dramatically reduce the enormous power consumption of AI applications.

The team of scientists at TU Braunschweig, Leibniz University Hannover, the Physikalisch-Technische Bundesanstalt (PTB) and Ostfalia University of Applied Sciences has a clear vision: to reduce the energy consumption of artificial intelligence to a minimum with a novel computer design. Instead of transistors, the data centre uses microscopic LEDs. “The LEDs give us the unique opportunity for massive parallel networking, comparable to the neural networks in the human brain. Our current platforms already have up to a quarter of a million LEDs, which can be matched by just as many sensors. In this way, we can optimise our computer design for typical AI applications such as pattern recognition,” says Professor Andreas Waag from TU Braunschweig, spokesperson of the NTC.

A team at the LENA Research Centre in Braunschweig has already successfully demonstrated how the innovative computer design works. Over the next five years, the project team aims to optimise this approach, expand the network and further develop key components.

To achieve this, the researchers are combining the strengths of two previously separate areas of microelectronics. On the one hand, there are silicon-based integrated circuits (CMOS), the backbone of chip production around the globe. These versatile circuits enable high technology to be implemented in the smallest of spaces for a wide variety of purposes. However, silicon chips cannot emit light for fundamental physical reasons. On the other hand, chips based on materials such as gallium nitride are highly luminous. The NTC aims to combine the advantages of both technologies in a single system. This hybrid integration will benefit not only the research consortium’s neuromorphic computers, but also a whole range of new applications in microelectronics.

“With BRIGHT, our researchers are impressively demonstrating how scientific progress can take responsibility for the future. The combination of light, microelectronics and neuromorphic thinking opens up a path to powerful AI that consumes significantly less energy and is therefore more sustainable. I would like to thank the state of Lower Saxony and the Volkswagen Foundation for their trust. Particular mention must be made of the vision and courage of the dedicated team in Braunschweig and Hanover in developing such a forward-looking topic on the basis of excellent research and collaboration. Together, we are laying the foundations today for technologies that will shape our society tomorrow,” says Angela Ittel, President of TU Braunschweig.

About Neuromorphic Computing

The global energy demand of data centres is growing four times faster than in all other sectors, driven by AI applications. According to the International Energy Agency (IEA) scenario, data centres will consume three per cent of global electricity – roughly double the current total consumption of Germany. This enormous energy hunger is primarily due to the simulation of neural networks. In order to recognise a pattern, for example, classic computers have to break down all data into 0s and 1s, complete as many training runs as possible with large data sets, and process them in a multi-stage process.

Neuromorphic computers, on the other hand, are designed to implement the required networks directly at the hardware level instead of simulating them with 0s and 1s. The NTC consortium’s neuromorphic system uses LEDs that enable parallel networking similar to neurons in the human brain. At the same time, the microscopic LEDs can be expected to consume minimal energy resources.

About the project

The state of Lower Saxony and the Volkswagen Foundation are funding the NTC BRIGHT project with around 15 million euros. The spokespersons for the consortium, Professor Andreas Waag (spokesperson) and Professor Bernhard Wicht (co-spokesperson), lead an interdisciplinary team of researchers at TU Braunschweig, Leibniz University Hannover, the Physikalisch-Technische Bundesanstalt (PTB) and Ostfalia University of Applied Sciences. Starting on 1 April 2026, the funding will run for five years until 2031.