iPronics Announces a SiPh Packaging and Assembly Manufacturing Line for Next Generation AI Optical Circuit Switches at Fabrinet

VALENCIA, Spain, March 17, 2026 (GLOBE NEWSWIRE) -- iPronics, the leader in programmable silicon photonics (SiPh) for AI datacenter networking, today announced a major expansion of its global manufacturing capability through a deepened partnership with Fabrinet (NYSE: FN), a world-class, Tier 1 provider of advanced optical and electronic manufacturing services.

Following early wins for ONE‑32 with the company’s first ten hyperscaler and AI‑cluster OEM customers, iPronics will ramp up production of SiPh‑based optical circuit switch (OCS) systems to meet demand for scalable, energy‑efficient interconnects in AI infrastructure.

iPronics is unveiling a dedicated manufacturing line for the packaging, assembly, testing, and qualification of its ONE Series SiPh OCS, enabling volume production from wafers to line cards that integrate into the customer’s rack. The new line is expected to be fully operational in the second quarter of 2026, supporting the company’s roadmap to serve hyperscaler, AI system integrator, and accelerator vendors.

A Scalable Manufacturing Platform for AI-Era Networking

As AI clusters grow toward hundreds of thousands of GPUs, traditional copper and electronic packet switching networks face hard limits in connectivity, latency, reliability, and power. Silicon photonics OCS — especially in scale-up architectures — has emerged as a critical enabler for more flexible and power efficient AI workloads.

iPronics’ partnership with Fabrinet extends an established collaboration on optical sub‑assemblies and introduces:

• Tier 1 assembly of OCS modules and 1U rack systems: Including flip-chip photonic integrated circuit (PIC) bonding, semiconductor optical amplifier (SOA) integration, thermal/mechanical qualification, and system-level testing.
• A new production line dedicated to iPronics OCS: Enabling end-to-end manufacturing of iPronics ONE Series OCS systems.
• Mass production readiness for scale-up AI clusters: Designed to meet the volume, reliability, and cost requirements of global hyperscalers.

Volume Ramp: From Pilot to Mass Production

The new manufacturing line enables iPronics to scale production rapidly to meet customer demand for SiPh OCS in training and inference AI clusters. Based on the company’s roadmap and customer engagements, the line is designed to support:

• Volume production of SiPh OCS modules.
• Fully automated optical alignment, packaging, and test infrastructure.
• End-to-end quality and reliability processes aligned with hyperscaler requirements.

This capacity supports iPronics’ anticipated growth in demand for OCS units, driven by deployments in scale-up and scale-out networks, within a market projected to reach multibillion dollar size by 2030.

For more information, contact press@ipronics.com

Leadership Statements

Christian Dupont, CEO of iPronics, said:
“AI datacenters are scaling beyond the limits of copper and electronic switching, Silicon photonics OCS is a strategic element in the infrastructures. Our partnership with Fabrinet provides confidence in our ability to deliver the manufacturing robustness, reliability, and high-volume scalability that Tier 1 customers expect. With this new production line, iPronics becomes the first company to offer a mass manufacturable, reliable, solid-state OCS solution built on a standard silicon photonics process.”

About iPronics

iPronics is a pioneer in the development of programmable photonic hardware for high-speed optical networks. Founded in 2019, iPronics is a spin-off of the iTEAM Photonic Research Laboratories of the Universitat Politècnica de València. Thanks to our patented photonic integration technology, we enable the design and deployment of ultra-fast, energy-efficient, and reconfigurable data infrastructures. Our solutions power next-generation artificial intelligence (AI), cloud computing, and high-performance computing ecosystems, providing unmatched flexibility, speed, and sustainability.

A photo accompanying this announcement is available at https://www.globenewswire.com/NewsRoom/AttachmentNg/44c0546b-6b14-4e78-a7d0-bf64ef4f310b