Nano electro-optomechanical programmable integrated circuits

Informazioni relative al progetto

NEUROPIC

ID dell’accordo di sovvenzione: 101098961

DOI

10.3030/101098961

Data della firma CE 15 Dicembre 2022

Data di avvio 1 Marzo 2023

Data di completamento 28 Febbraio 2027

Finanziato da

• The European Innovation Council (EIC)

Costo totale

€ 2 999 922,74

Contributo UE

€ 2 999 921,50

Coordinato da

AGENCIA ESTATAL CONSEJO SUPERIOR DE INVESTIGACIONES CIENTIFICAS
Spain

Periodic Reporting for period 2 - NEUROPIC (Nano electro-optomechanical programmable integrated circuits)

Periodo di rendicontazione: 2024-03-01 al 2025-08-31

Sintesi del contesto e degli obiettivi generali del progetto

Modern society faces unprecedented demand for data processing driven by artificial intelligence, autonomous systems, data centers, and emerging quantum technologies. These applications require computing systems that are not only faster but also dramatically more energy-efficient than current technologies. Photonic integrated circuits—chips that process information using light rather than electrons—offer a promising path forward, but existing approaches for making these circuits programmable often suffer from high power consumption and thermal management challenges.
NeuroPIC addresses this fundamental challenge by exploring alternative approaches to programmable photonics that offer the potential for significantly reduced power consumption compared to conventional technologies while providing compact footprints and potentially faster operation. Recent technological advances now make it feasible to explore these concepts at larger scales, opening possibilities for programmable photonic processors that were previously impractical.

The project pursues several interconnected technical objectives spanning advanced nanofabrication, scalable programmable photonic platforms, high-density optical interconnection technologies, and neuromorphic computing implementations. These objectives aim to demonstrate that programmable photonic circuits can perform complex information processing tasks with improved energy efficiency compared to existing solutions. A key scientific goal is exploring neuromorphic computing—brain-inspired information processing—using photonic systems with nonlinear dynamics. This research investigates fundamental questions about how physical complexity and nonlinearity contribute to computation, potentially revealing new principles for next-generation computing architectures beyond conventional digital approaches.

Lavoro eseguito dall’inizio del progetto fino alla fine del periodo coperto dalla relazione e principali risultati finora ottenuti

The project has progressed across nanofabrication, programmable photonics, optical interconnections, and neuromorphic computing.

Advanced Silicon Nanofabrication
Nanofabrication processes for nanoscale silicon structures were developed and optimized. Challenges revealed fundamental insights into nanoscale patterning physics. Theoretical frameworks were developed to understand fabrication trade-offs.

Programmable Photonic Platform Development
Photonic components for programmable circuits were developed. Multiple architectural approaches for large-scale networks were designed. Advanced fabrication workflows were established, providing foundational capabilities for scaling.

Optical Interconnection Technologies
A high-density optical interconnect demonstrator was developed and characterized. Custom fiber array technologies were co-developed. Packaging designs for complex photonic systems were completed and validated.

Neuromorphic Computing Implementation
Theoretical frameworks for nonlinear dynamics in photonic structures were established, demonstrating information processing capabilities. An experimental platform was constructed and validated. Device designs were developed and demonstrator concepts defined for further validation.

Progressi oltre lo stato dell’arte e potenziale impatto previsto (incluso l’impatto socioeconomico e le implicazioni sociali più ampie del progetto fino ad ora)

The project has achieved results that advance beyond current capabilities in programmable photonics, nanofabrication, and neuromorphic computing.
In nanofabrication, the consortium has pushed silicon lithography boundaries to achieve feature sizes and structural control exceeding conventional capabilities. Novel insights into ultra-high-resolution patterning physics have been uncovered. Some advances remain confidential for intellectual property protection.

Large-scale programmable photonic networks demonstrate potential advantages in power efficiency. Novel architectural approaches enable scaling to thousands of programmable nodes, with error correction methodologies established for practical deployment.
High-density optical interconnect demonstrations achieved channel densities substantially exceeding conventional standards. Custom fiber arrays with reduced pitch dimensions demonstrate advanced interconnection manufacturability.
In neuromorphic photonic computing, theoretical frameworks and experimental platforms exploring native nonlinearities in photonic nanostructures have been established, demonstrating functional computing capabilities using physical dynamics.