New chip opportunity Researcher works on neuromorphic chips for the industry

Related Vendors

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

Wuxi InfiMotion Technology Co., Ltd.

Neuromorphic chips, which process information like our brain, are the forte of physicist Prof. Dr. Heidemarie Krüger. What they can bring to the industry is revealed in this contribution ...

Together with her team, Krüger is developing components based on so-called memristors, which are claimed to set new standards in energy efficiency and computing power. These real-time capable and resource-efficient systems could support self-driving cars and industrial plants, for example. "Our goal is to use the brain as a model to create something that can logically understand complex decisions with minimal energy consumption," as the scientist points out. Accordingly, the neuromorphic chip is based on memristors. These are components that work similarly to synapses in the brain. They not only store information but can also process it simultaneously, it is stated more precisely. While conventional computers constantly exchange data between memory and processor, this type of chip operates locally, significantly reducing energy losses and allowing for fast, decentralized data analysis. A key difference is the ability of memristors to process continuous intermediate states—not just 0 and 1—but also values in between. This flexible data processing opens up new possibilities for algorithms that mimic neural networks. Applications range from predictive maintenance of machinery to real-time analysis in safety-critical areas such as autonomous driving.

Neuromorphic chips react highly sensitively to changes in machines

The architecture of memristors allows data to be processed directly at the source, which is a key component for so-called edge computing, where data does not need to be transferred to central cloud systems. This increases security and independence, as sensitive data remains local. Specifically, in the context of industrial sensor technology, this can be a major advantage to detect early signs of wear and prevent system failures. In initial pilot projects, Krüger's team is already testing the technology under real conditions with the Technical University Bergakademie Freiberg (Germany). It has been shown that the neuromorphic chip can reliably detect even the smallest changes and accurately predict wear patterns. While classical processors require more and more transistors to handle the growing data flood, traditional chip designs are slowly reaching their physical and energy limits. Neuromorphic approaches, however, combine memory and processing units, reducing energy needs and significantly expanding the potential for AI systems, as Krüger explains. The research aims not only to analyze data sets but also to learn to recognize patterns to respond flexibly to new situations without the need for constant connection to external data centers. Incidentally, Krüger's current prototype operates with 32 memristors. The next version is expected to have 200.