From Research Lab to Startup: How a Neuromorphic Chip Could Benefit Industry
This real-time and resource-efficient technology could support applications such industrial plants
Neuromorphic chips that process information like the human brain – this is the goal of physicist Heidemarie Krüger and her Dresden-based startup “Techifab”. The researcher from the Leibniz Institute of Photonic Technology and the Friedrich Schiller University of Jena is developing a technology that processes and stores data directly at the point of origin, eliminating the need for energy-intensive data transfers between processor and memory. Together with her team, Krüger is working on memristor-based components that will set new standards in energy efficiency and computing power. This real-time and resource-efficient technology could support applications such as autonomous vehicles and industrial plants. “Our goal is to use the brain as a model to create a technology that can make complex, logical decisions with minimal energy consumption,” says Heidemarie Krüger.
The Core Innovation: Memristors With Memory and Learning Capabilities
The neuromorphic chip is based on memristors – components that function similarly to synapses in the brain. They can not only store information, but also process it simultaneously. While conventional computers continuously exchange data between memory and processor, this technology works locally. This significantly reduces power dissipation and enables fast, decentralized data analysis.
A key difference is the ability of memristors to process continuous intermediate states – not just ‘0’ and ‘1’, but values in between,” explains Krüger. This flexible data processing opens up new possibilities for algorithms that simulate neural networks. Potential applications range from predictive machine maintenance to real-time analysis in safety-critical areas such as autonomous driving.
From Laboratory Discovery to Industrial Application
The journey to this innovation began with a serendipitous discovery in the lab in 2011. During a materials analysis, Krüger’s team observed a characteristic “loop” curve – a signature behavior of a memristor with hysteretic memristor resistance. This property allows the device to “remember” past computations and perform complex calculations directly. This discovery led to the idea of developing artificial synapses using a combination of bismuth and iron oxide. To turn these artificial synapses into a functional chip, the startup received millions of euros in funding from Germany’s Federal Agency for Disruptive Innovation.
“We’ve shown that these artificial synapses can efficiently handle complex computational tasks such as matrix multiplication,” Krüger says. Such calculations form the basis for training many AI applications and image processing algorithms. In January 2025, the news magazine “Der Spiegel” reported on how Krüger’s technology could set new standards in energy-efficient computing.
Technology With Potential for Edge Computing
The architecture of memristors allows data to be processed directly at the source – a key component for so-called edge computing, where data does not need to be transferred to central cloud systems. “This means increased security and independence, since sensitive data remains local,” Krüger points out. This could be a significant advantage in industrial sensor systems, for example, to detect early signs of wear and prevent system failures.
In initial pilot projects, Krüger’s team is already testing the technology under real-life conditions in collaboration with the Technical University of Freiberg. The tests have shown that the neuromorphic chip can reliably detect even the smallest changes and accurately predict wear patterns.
A Sustainable Path to Energy-Efficient AI Systems
While conventional processors require more and more transistors to handle the growing flood of data, traditional chip designs are reaching physical and energy limits.Neuromorphic approaches combine memory and processing units, reducing energy consumption and significantly expanding the potential for AI systems.
“Our goal is not only to analyze data sets, but also to learn, recognize patterns, and react flexibly to new situations without being constantly connected to external data centers,” says Krüger. This technology could help make data centers more energy efficient and enable AI applications to be developed with significantly fewer resources.
Krüger’s current prototype has 32 memristors. In the next phase of development, this number is expected to increase to over 200 to model complex neural networks and enable new applications in autonomous systems.
Original publication
Yao Shuai, Shengqiang Zhou, Danilo Bürger, Manfred Helm, Heidemarie Schmidt; "Nonvolatile bipolar resistive switching in Au/BiFeO3/Pt"; Journal of Applied Physics, Volume 109, 2011-6-29
Heidemarie Schmidt; "Prospects for memristors with hysteretic memristance as so-far missing core hardware element for transfer-less data computing and storage"; Journal of Applied Physics, Volume 135, 2024-5-29
Other news from the department business & finance
Most read news
More news from our other portals
Something is happening in the chemical industry ...
This is what true pioneering spirit looks like: Plenty of innovative start-ups are bringing fresh ideas, lifeblood and entrepreneurial spirit to change tomorrow's world for the better. Immerse yourself in the world of these young companies and take the opportunity to get in touch with the founders.