The 2DNeuralVision project is redefining how we think about computing by merging neuromorphic principles with photonic architectures. At the International School of Physical Computing (ISOPC25) in Erice, Italy, our team showcased groundbreaking research that pushes the boundaries of hybrid computing systems, combining quantum, neuromorphic, and classical paradigms for next-generation information processing.

Why Neuromorphic Photonics Matters

As traditional electronics approach physical and energy limits, photonic computing offers a revolutionary alternative. By using light instead of electrons, photonic systems deliver:

• Ultra-fast data processing with minimal latency.
• Energy efficiency far beyond conventional semiconductor technologies.
• Scalability for complex machine learning and neuromorphic tasks.

Neuromorphic photonics takes this further by mimicking brain-like architectures, enabling parallel processing, adaptability, and integration with quantum technologies.

Our Contribution: Scalable Photonic Architectures

During the workshop, Dr. Shabnam Taheriniya presented our latest research in the talk: “Designing Scalable Neuromorphic Photonics with Electron Microscopy and Beam-Assisted Trimming”

Key innovations include:

• Electron Microscopy for Feedback: High-resolution imaging provides critical insights into device morphology, enabling precise control over photonic circuits.
• Beam-Assisted Trimming: A post-fabrication technique that fine-tunes optical properties for optimal performance and scalability.
• Hybrid Computing Paradigms: Integrating quantum, neuromorphic, and classical approaches to create architectures capable of handling next-generation AI workloads.

Global Collaboration and Future Outlook

The event highlighted the importance of international collaboration in shaping the future of computing. Prof. Wolfram Pernice played a key role in organizing this workshop, which will now be held annually, creating a platform for researchers to exchange ideas and accelerate innovation in photonic computing and machine learning.

Learn more about the event: https://icopc25.kip.uni-heidelberg.de/