Imagine a computer chip that works like a human brain, processing information using magnetic ripples instead of moving electricity through wires. By using tiny magnetic discs, it can recognize patterns and predict trends without needing to move data around, which saves a lot of energy. It's like a musical instrument where the vibrations themselves do the math.
Current AI computing at the edge is too energy-intensive and requires heavy data pre-processing. There is a critical need for hardware that can process complex signals, like radar data, with minimal power consumption.
A hardware solution for brain-inspired computing using nanoscale magnetic materials and spin-wave interactions. This includes optimized NiFe samples and a proof-of-concept device for pattern recognition.
If you are a vehicle software provider dealing with the high energy cost of processing radar data for gesture recognition — this project developed a magnon reservoir that enables pattern recognition with minimal pre-processing. This allows for faster, lower-power feature recognition in autonomous driving scenarios.
If you are a network equipment manufacturer dealing with the need for THz frequency operation — this project developed research into synthetic and pure antiferromagnetic materials. This provides the groundwork for 6G compatibility and high-speed signal processing.
If you are a hardware designer dealing with the power constraints of edge computing — this project developed a hardware solution using nanoscale magnetic materials. This creates an energy-efficient computing scheme for AI tasks like time series prediction.
Based on available project data, specific unit costs or pricing models are not provided; the project focuses on developing the proof-of-concept and scalable processes.
The project aims to develop a proof-of-concept device using industrially compatible processes to ensure scalability, targeting TRL 4.
Based on available project data, specific licensing terms are not listed, but the project involves 2 industrial partners and 3 research centers, suggesting a collaborative IP environment.
The technology acts as a hardware reservoir for computing, specifically designed for tasks like pattern recognition and time series prediction with minimal input pre-processing.
The project period runs from 2022-10-01 to 2026-09-30, with objectives ranging from TRL 1 to TRL 5.
The consortium consists of 8 partners across 4 countries (DE, FR, NL, PT). It has a balanced mix of 3 universities, 3 research centers, and 2 industrial partners, resulting in an industry ratio of 25%. This structure suggests a strong bridge between fundamental physics research and industrial application.
Contact HELMHOLTZ-ZENTRUM DRESDEN-ROSSENDORF EV
Contact us to explore licensing opportunities for spintronic AI hardware.
