Imagine a computer chip that uses light instead of just electricity to think, making it incredibly fast and energy-efficient. It uses special hybrid particles that act like a bridge between light and matter to process information. This allows AI tasks to happen almost instantly without needing a massive power plant to run them.
Current AI computation relies on software simulations and electronic hardware that consume too much power and are too slow for real-time, on-device processing in small gadgets.
A scalable fabrication method for CsPbBr3 microwire waveguides and a polariton-based Toffoli-like logic gate for neural network acceleration.
If you are a clinical diagnostic lab dealing with slow genomic analysis of microarrays — this project developed a polariton-based accelerator that drastically reduces the time needed for biomarker pattern detection. This speeds up drug development and patient prognostics.
If you are a hardware maker dealing with high power consumption in sound and image recognition — this project developed neuromorphic hardware that processes these tasks with reduced power consumption. It enables complex AI in small devices that cannot connect to large remote servers.
If you are a chip manufacturer dealing with the limitations of standard electronics for AI — this project developed a scalable method for fabricating CsPbBr3 microwire waveguides. These support room-temperature operation, making them compatible with standard electronics.
Based on available project data, specific pricing or cost structures are not provided as the project is currently in the research and development phase.
The project has developed a scalable method for fabricating microwire waveguides using a microfluidic-assisted process with PDMS templates, which is described as promising for on-chip photonics.
Based on available project data, there are no specific details regarding patents or licensing agreements mentioned in the report.
The project specifically aims to ensure compatibility with standard electronics to allow the polariton-based neural network to function as a hardware accelerator.
The project period runs from 2024-02-01 to 2028-01-31, suggesting that the technology is still in the development and demonstration stage.
The consortium is well-balanced for a deep-tech project, consisting of 8 partners across 4 countries. With a 25% industry ratio (2 industrial partners, including 1 SME), there is a clear bridge between the 3 universities and 3 research institutes, ensuring that the theoretical physics of polaritons is translated into practical hardware fabrication.
Contact the research office at Uniwersytet Warszawski
Contact us to identify potential licensing opportunities for polariton-based AI hardware.
