A new technology links two ultra-small magnetic oscillators so they “talk” to each other via magnetic fields in tiny waveguides. By splitting and feeding back parts of their own signals, the oscillators lock into step improving signal strength and clarity. This approach works over distances of millimeters to meters, lets engineers fine-tune both the timing and power of the link, and could pave the way for ultra-efficient radio devices and brain-inspired computing hardware.
Figure1. Two spin torque nano-oscillators are synchronized using Oersted field coupling. Each MTJ has a CPW on top, DC-biased via bias-tees. RF outputs are split – one part amplified and fed to the other CPW for coupling, the other combined and analyzed on a spectrum analyzer.
Modern electronic systems, especially those used in communication and advanced computing, require highly accurate and energy-efficient signal generators. Traditional methods often face challenges in maintaining stability and power efficiency, especially at very small scales. There is also a growing need for hardware that can mimic how the brain works for future AI applications.
- On-Chip Feedback Loop: This product uses on-chip waveguides to route parts of each oscillator’s output back into its partner, creating a self-contained feedback loop.
- Variable Coupling Strength: This product allows independent control of link strength via adjustable amplifiers, so synchronization range can be widened or narrowed at will.
- Phase-Shift Tuning: This product offers phase-shifting capability, letting users set the exact timing offset between oscillators for optimal performance.
- Extended On-Chip Range: This product works over long on-chip distances (up to meters), enabling scalable arrays without performance loss.
- Crosstalk Elimination: This product’s design prevents unwanted electrical crosstalk by using isolators, ensuring pure magnetic coupling only.
In the lab, researchers built a complete physical setup on a silicon wafer. Tiny magnetic tunnel junctions (about 100 nm across) were capped by 2 μm-wide coplanar waveguides. Each junction was powered by direct current, and its radio-frequency output was split: one part went to measurement equipment, the other was boosted by an amplifier and fed into its partner’s waveguide. Phase shifters and isolators were added to fine-tune timing and block stray signals. Experiments varied magnetic fields, currents, amplifier gains, and phase delays to map out when and how the oscillators locked together.
A working prototype was demonstrated using advanced measurement techniques. It includes tiny magnetic components, signal boosters, and phase adjusters to manage how signals are sent and received between the two parts.
4
By enabling tiny oscillators to synchronize reliably, this technology can lead to more energy-efficient radios, better wireless communication devices, and compact signal generators for satellites or IoT sensors. Its ability to mimic neuron-like coupling could accelerate the development of hardware that thinks more like a brain—powering next-generation AI with far lower energy use. In turn, smarter and greener electronics benefit everyday life, from longer-lasting smartphones to more responsive autonomous vehicles and advanced medical imaging systems.
- Advanced radio-frequency signal generators
- Neuromorphic (brain-inspired) computing hardware
- Wireless communications (5G/6G infrastructure)
- Internet of Things (miniature sensor nodes)
- Satellite and aerospace electronics
- Defence and radar systems
- On-chip clock distribution networks
Geography of IP
Type of IP
201921018149
542538