Intelligent Manufacturing for AI at Scale

10:40 am - 11:10 am

The explosive growth of AI and high-performance computing is pushing semiconductor architectures toward ever tighter integration of photonics, electronics, packaging, and test. In this context, Co-Packaged Optics (CPO) is emerging as a critical convergence point for next-generation bandwidth, power efficiency, and system performance. Yet the key challenge is no longer proving that these technologies work - it is turning them into scalable, repeatable, and efficient manufacturing systems.

This talk argues that smart manufacturing is the missing link between innovation and deployment at AI infrastructure scale. Sustainable AI does not start in the data center - it starts in manufacturing, where yield, throughput, and process efficiency determine whether advanced heterogeneous integration can move from promising technology to industrial reality.

Using CPO as a representative example, the presentation highlights three critical pillars for scalable photonic manufacturing: wafer-level testing, fully automated fiber optic cable assembly, and high-precision active optical alignment. Together, these capabilities enable the transition from mature process building blocks - such as fiber-chip coupling and active alignment - to high-volume manufacturing under real-world variability.

The core message is clear: scaling is not a machine problem, but a system problem. It requires the convergence of automation, modular production architectures, and real-time data-driven process control. Embedded intelligence - through machine data, KPI tracking, and adaptive optimization - makes AI part of the manufacturing process itself.

Ultimately, AI creates demand - but intelligent manufacturing determines whether heterogeneous integration can scale to meet it.

Key Technologies Covered

Smart manufacturing for AI
Wafer level testing
Active Assembly
Heterogeneous Integration
Automated fiber optical cable manufacturing
Co-packaged optics
Intelligent manufacturing for AI
Automation & Data driven production

Featured Speakers

Dr. Moritz Seyfried

Director Research & Development, ficonTEC Service GmbH

Moritz holds a Master’s degree from Heriot-Watt University in Edinburgh, UK, a Diploma degree from the Karlsruhe Institute of Technology (KIT), Germany, as well as a PhD from Bremen University, Germany – all in the field of semiconductor optics. He spent most of the following ten years working in the field of organic and inorganic semiconductor optics, ranging from basic research of light-matter interaction phenomenon to photonics product assembly. Moritz joined ficonTEC Service in 2012 and since 2014 is the Manager of ficonTEC R&D, located within ficonTEC HQ in Achim, Germany.