Table of contents:
- IT hardware consumes a third of the electrical energy
- Greater energy efficiency with neuromorphic electronic systems
- Simulate information processing between nerve cells
- Intelligent materials for neuromorphic electronics
- Material analysis for customized electronic components
Video: Researchers Are Studying Materials For Biologically Inspired Electronics
2023 Author: Hannah Pearcy | [email protected]. Last modified: 2023-05-24 11:12
The digital revolution and the increasing use of artificial intelligence are changing technologies and our society. However, this goes hand in hand with a steadily growing energy requirement and ever higher carbon dioxide emissions.
IT hardware consumes a third of the electrical energy
According to the TU, the hardware that is used in IT applications around the world is already consuming a third of the total electrical energy produced worldwide - and the trend is rising: Scientific projections predict that the entire global electrical energy production will take place in around 15 years will no longer be sufficient to cover the power requirements of the IT hardware.
Greater energy efficiency with neuromorphic electronic systems
Scientists at the TU Ilmenau therefore rely on so-called neuromorphic electronic systems that promise far higher energy efficiency than today's electronics. In these electronic systems inspired by biology, memristive materials are the central building block. With their "memory", these intelligent materials, learning and memory processes of biological systems enable technical replication. The word memristiv is composed of the English "memory" for memory and "resistor" for electrical resistance.
Simulate information processing between nerve cells
Such so-called memristive materials can be used to produce electronic components that simulate the biological information processing between nerve cells. This should make it possible to implement energy-efficient electronics that make an important contribution to reducing the global carbon dioxide burden. The joint project "Memristive Materials for Neuromorphic Electronics (Memwerk)" is funded by the Carl Zeiss Foundation as part of the "Breakthroughs" program with EUR 4.5 million for five years.
Intelligent materials for neuromorphic electronics
In the collaborative project "Memristive Materials for Neuromorphic Electronics (Memwerk)", scientists from materials science, computer science and electrical engineering and information technology are researching how the information processing of biological systems can be converted into energy-efficient technical systems.
The aim of the research project is tailor-made intelligent memristive materials for neuromorphic electronics. Prof. Ziegler, scientific director of the project, is confident that memristive components will enable highly energy-efficient systems to be developed in the future: “We will be able to technically reproduce the biological paradigms of information processing, learning and memory formation more precisely than ever before and create completely new opportunities for information technology."
Material analysis for customized electronic components
In theoretical and experimental work, a 26-member research group from six specialist areas of the TU Ilmenau will carry out comprehensive material analyzes, design innovative neural network structures, develop memristive materials and model and manufacture electronic components from these materials in order to implement energy-efficient neuromorphic circuits.
The scientists will also develop a digital mapping system for memristive materials that directly relates the material properties and technological parameters of material synthesis and component development to the characteristics and performance parameters of neuromorphic circuits. From this approach, the scientists hope to tailor materials for neuromorphic electronics.
