Neuropic, an EIC Pathfinder under the coordination of ICMM-CSIC that aims to develop neuromorphic computing, held its third annual meeting in the first week of September in Braga (Portugal). Hosted by the International Iberian Nanotechnology Laboratory (INL), the annual event allowed the consortium to meet in person to discuss the advances and address the project's challenges, 18 months after its launch last March 2023.
Pedro David García, the project's coordinator and researcher at the ICMM, explains that NEUROPIC is approaching its second year and the meeting in Braga allowed the team "to review its current state and discuss how to face its progress in the coming months". Up to now, the consortium has achieved "significant progress" on the fabrication aspects (Nature 624, 57) of the nanoscale systems "obtaining very good performance of the programable linear network at small scale", he says. "The consortium is also making promising progress on the computation exploiting the different nonlinearities of the system", he adds.
In the NEUROPIC project, European researchers, innovators, and companies in nanotechnology, photonics, advanced manufacturing, and complex systems join forces to explore programmability using nano-electro-mechanical technology and its applications in neuromorphic computing. The teams, from five European countries (Spain, Denmark, Germany, Ireland and Portugal), are imitating in electronic devices how the human brain works through a complex system of interconnected neurons. It is known as neuromorphic computing, an alternative way of processing information different from standard computers or mobile phones. They will also do so with a new and sustainable approach: "We propose using thermal energy from the system to create this type of artificial neurons," says García.
"EUROPIC is an EIC Pathfinder Open 2023-funded project that brings together 5 European countries (ES, DK, DE, IE and PT) to build novel programmable photonic chip architecture that could have important implications for various fields, including in data centres, autonomous vehicles and quantum information processing", explains García.
The 6 project partners are currently exploring novel methods of programming light signals on chips with a high level of flexibility. The team will integrate nanoelectromechanical technologies into the silicon photonics platform that are characterised by ultralow power consumption, small footprint and potentially faster operation compared to existing approaches. The ultimate goal is to build a network of independent and fully controllable nodes for neuromorphic computing.