Inspired by nature
They could be the key to addressing the energy demands of Artificial Intelligence (AI): neuro-inspired computer chips modeled after the human brain. Researchers at the Forschungszentrum Jülich from the Helmholtz research area for Information and RWTH Aachen University are significantly advancing the technology in the NEUROTEC project and the Future Cluster NeuroSys. At the Jülich-Aachen Neuromorphic Computing Day, scientists, in collaboration with partners from high-tech companies and start-ups, provided the roughly 200 guests from academia, business, and politics, including BMBF State Secretary Sabine Döring, with insights into the current state of developments. (Source: Forschungszentrum Jülich – Press Releases)
Please note: The text has been translated using a machine translation tool and may contain inaccuracies.
“We aim to promote structural change in the Rhenish district through research and innovation. In this way, we ensure that value creation and employment in the regions are maintained and strengthened. The NEUROTEC project demonstrates how this can be achieved: Building a strong network of science and industry pushes forward the microelectronics of the future, accelerating its transfer to businesses. Through NEUROTEC and the Future Cluster NeuroSys, the Greater Aachen area aims to become one of the leading sites for European AI hardware. Because neuromorphic chips, as developed in NEUROTEC and NeuroSys, can significantly reduce AI’s resource requirements in the future. However, a successful structural change can only succeed if we also manage the transfer of scientific results into practice. NEUROTEC and NeuroSys offer optimal conditions for this, thus establishing a new, innovative AI hub in the Rhenish district,” said the State Secretary in the Federal Ministry of Education and Research (BMBF), Prof. Sabine Döring, at the start of the networking event in the large auditorium of the Forschungszentrum Jülich.
Data centers in Germany and worldwide have required increasing amounts of electricity for years. The use of AI is considered one of the drivers of this development. ChatGPT and similar platforms exhibit increasingly human-like capabilities. However, the computers running these applications function quite differently from a biological brain. The difference is also evident in energy consumption: Supercomputers require as much electricity to train artificial neural networks as a small city. In contrast, the human brain consumes no more energy than a 25-watt light bulb.
“Neuromorphic systems, which emulate the functionality of the brain, promise to train and operate such AI processes much more efficiently – by several orders of magnitude – than is possible with conventional digital computers,” explains Prof. Astrid Lambrecht, Chairwoman of the Forschungszentrum Jülich. “This technology represents a genuine key innovation. Neuromorphic computing is an interdisciplinary research field par excellence, which could not unfold without the networking of neuroscience, physics, electronics, computer science, and materials science, as practiced here at the Forschungszentrum Jülich and in collaboration with our partners. Potential applications range from intelligent implants in medicine, pattern recognition in smartphones and other mobile devices, to the use of AI in industry, agriculture, and autonomous driving.”
The Forschungszentrum Jülich and RWTH Aachen University are considered global leaders in researching memristive materials and components. These are a central element for constructing a neuromorphic computer. Memristive components can store and process information in parallel as artificial synapses in neural networks.
However, NEUROTEC goes beyond pure research. In collaboration with high-tech companies in the region, such as equipment manufacturer AIXTRON, measurement technology specialist aixACCT Systems GmbH, and nanotechnology companies AMO GmbH and SURFACE systems+technology GmbH & Co. KG, a new, comprehensive basic technology for neuromorphic AI is emerging, and technological know-how flows directly from research to – primarily regional – businesses. The joint project entered its second phase in 2021 and is funded by the BMBF for five years with 36 million euros from funds for structural change.
“The first results will already be on the market in the next generation of automotive electronics. This involves energy-saving and fast ReRAM memories from Infineon and TSMC that can store information permanently without continuous power supply. These have been optimized using test and programming methods from NEUROTEC,” explains NEUROTEC coordinator Prof. Rainer Waser from Forschungszentrum Jülich and RWTH Aachen University.
The actual goal of the project, the realization of the first neuromorphic demo chips for practical use, is also increasingly within reach. “We have extensively experimented with chips based on 180-nm semiconductor technology and made many smaller, significant advances. This involves, for example, the question of how reliably these novel components can be produced using CMOS technology, or whether they still function flawlessly even after many switching cycles,” reports Prof. Rainer Waser.
The first designs based on the more modern 28-nm technology went into production in Taiwan at the end of May. After further processing and equipping with memristive elements from Jülich-Aachen research, they will soon be tested with real applications, such as methods for route optimization, computer virus searches in the data stream, or artificial neural networks on a hardware level for deep learning.
Complementary to this, researchers from RWTH Aachen University and Forschungszentrum Jülich in the Future Cluster NeuroSys, in collaboration with companies in the region, are working to tap further market and application potentials.
“Neuromorphic Computing will open up numerous application opportunities for energy-efficient AI applications in speech, image, and video processing, or medicine. The RWTH thinks big and offers its extensive experience in translating research results. Together with associations like the IHK Aachen or the Zukunftsagentur Rhenish District, various scenarios for the further development of our innovation ecosystem are considered, up to the establishment of a factory for computer chips ‘made in NRW’,” explains Prof. Ulrich Rüdiger, Rector of RWTH Aachen University.
The Aachen and Jülich region has so much to offer. In addition to the renowned research institutions, high-tech companies such as AIXTRON, aixACCT, Surface, Sympuls, X-Fab, and ELMOS are located there, active in the strategically essential field of the semiconductor industry. There are also start-up initiatives like NEUREKA (RWTH) and AiML (FZJ), both funded by the SPRIND agency as part of the New Computing Concepts Challenge, or the start-up Black Semiconductor (AMO), which is involved in the planned second IPCEI for microelectronics and communication technologies.
“With our research, we span the entire value chain, including sociological and ethical issues that disruptive technologies always raise. As a result, opportunities arise at many technological levels for regional companies and start-ups to accelerate structural change. We already see the first effects on the job market and corporate investments today. In the long term, semiconductor production in the region would ideally complement the existing, comprehensive, and internationally recognized expertise and infrastructure, and significantly increase the attractiveness of the region for the excellently trained minds from the research center and RWTH,” says NeuroSys coordinator Prof. Max Lemme of RWTH Aachen University and Managing Director of AMO GmbH.
With the Jülich Supercomputing Centre (JSC) at the Forschungszentrum Jülich, the leading international institution for the operation and construction of supercomputers, there is also the necessary infrastructure available to put new developments to the test immediately.
It’s important to remember that the possibilities and potential of neuromorphic computing are still in their infancy. The developments at Forschungszentrum Jülich and RWTH Aachen University underline the significance and future prospects of this groundbreaking technology in the global AI market.
FZJ/T. Schlößer, 30.08.2023
The original press release can be found at:
Von der Natur inspiriert (only in german)
Localization in the Helmholtz Research Field Information:
Helmholtz Research Field Information, Program 2: Natural, Artificial and Cognitive Information Processing, Topic 3: Neuromorphic Computing and Network Dynamics
Contact:
Prof. Dr. Rainer Waser
Coordinator NEUROTEC, Director of the Peter Grünberg Institute for Electronic Materials (PGI-7)
Forschungszentrum Jülich
Phone: +49 2461 61-5811
E-Mail: r.waser@fz-juelich.de
Prof. Dr. Ing. Max C. Lemme
Coordinator NeuroSys, Chair for Electronic Components
RWTH Aachen
Phone: +49 241 80-20280
E-Mail: neurosys@eld.rwth-aachen.de
Contact for this Press Release:
Tobias Schlößer
Press Officer
Forschungszentrum Jülich
Phone: +49 2461 61-4771
E-Mail: t.schloesser@fz-juelich.de
