Why Urgency Has Become the New Reality for Integrated Photonics

By Dr Abdul Rahim, Ecosystem Manager, PhotonDelta

When I look at the integrated photonics landscape today, two words define what I see: scaling and urgency.  Technical performance and emerging applications remain fundamental, but AI has added a new dimension to the conversation.

As AI systems grow, moving data between accelerators, switches, memory, and distributed compute sources is becoming a critical constraint, increasing demand for the bandwidth density and energy efficiency that integrated photonics can provide.

As high-speed electrical interconnects encounter growing constraints in reach, signal integrity, and energy consumption, optical communication is moving progressively towards co-packaged optics and optical-I/O architectures.

The market is signalling urgency

One of the clearest indicators is the investment landscape itself. Decisions that might previously have unfolded over many years are now being compressed into much shorter timescales. I recently reviewed around 90 major transactions and ecosystem developments between 2024 and 2026. From those, I was able to verify disclosed values for 49 major transactions, which I grouped into three categories: mergers and acquisitions, funding rounds, and strategic partnerships or investments.

The analysis covered both demand-side markets, including AI infrastructure, optical networking, quantum photonics, and sensing, and the enabling layers of the supply chain, including foundries, packaging, and testing. Looking across the full ecosystem provides a clearer picture of where organisations believe long-term value will be created.

The figures themselves are significant, representing more than $18 billion in disclosed activity. More revealing, however, is where that investment is being directed. If coherent optical transport is included as AI-enabling infrastructure, around 98% of disclosed M&A value and 100% of disclosed strategic investment value is AI-related. Funding rounds present a more balanced picture, with approximately 60% linked to AI and 40% supporting broader integrated photonics applications.

Together, these figures suggest that AI infrastructure is the dominant catalyst driving investment today. At the same time, sustained funding across quantum technologies, sensing, imaging, healthcare, and advanced computing demonstrates that integrated photonics is developing multiple long-term demand pathways beyond AI.

Perhaps the clearest signal of urgency is that companies increasingly want to control, or establish, much tighter coordination across critical parts of the stack. DSPs, SerDes, lasers, photonic integrated circuits, packaging, transceivers, optical engines, and system integration are all becoming part of broader strategic portfolios.

The reasons are technical as well as strategic. At these data rates and integration densities, performance is determined not only by individual components but also by the interfaces between them. DSP and SerDes architectures, electrical drivers, modulators and detectors, laser sources, fibre coupling, thermal management, and packaging increasingly need to be designed as a system. Loss, electrical parasitics, heat, and assembly tolerances at one layer can affect the performance and power budget of the entire optical link.

The objective is to secure manufacturing capacity, reduce supply-chain dependency, improve power efficiency and latency, and co-optimise complete optical links for AI-scale systems. This also places greater emphasis on manufacturability, as repeatable processes, yield, testability, and scalable optical packaging increasingly determine whether promising architectures can move from demonstration to high-volume deployment.

To me, this represents a structural shift in the industry, with companies positioning themselves for future demand rather than simply responding to markets that already exist.

AI is accelerating every part of the ecosystem

AI-scale deployment is forcing integrated photonics to mature simultaneously across design, manufacturing, packaging, and test. The challenge is no longer only to demonstrate high-performance photonic devices, but to manufacture them reliably, assemble them efficiently, and test them at a cost and throughput compatible with large-scale deployment.

For design teams, this means moving beyond isolated photonic components towards electro-optic co-design, more accurate compact models, process-corner analysis, and system-level simulation. For foundries, it requires tighter process control, higher repeatability, and manufacturing platforms capable of supporting higher-volume production. Packaging must combine precise optical alignment with high-density electrical interconnects and effective thermal management, while testing needs to become more automated, with earlier wafer- and die-level screening to avoid carrying defective components into costly assembly. At scale, yield, test time, assembly throughput, and process variability become first-order commercial constraints rather than simply manufacturing details.

This is also changing when different parts of the ecosystem become involved. Foundry, packaging, and test considerations increasingly need to be addressed during architecture and design decisions rather than after a photonic device has been developed. Hyperscalers, semiconductor manufacturers, packaging specialists, foundries, system integrators, and test-equipment suppliers are therefore becoming involved earlier in the development cycle, driving closer co-development and earlier investment in manufacturing capability.

Europe has the foundations to respond at scale

Against that backdrop, Europe is in a stronger position than is sometimes recognised. Over many years, sustained investment has created a broad integrated photonics ecosystem spanning research, design enablement, foundries, manufacturing, packaging, testing, and equipment suppliers.

Importantly, that ecosystem is also technologically diverse. Europe has established capabilities in silicon photonics, InP, and SiN, alongside emerging material platforms and heterogeneous integration. These are complemented by expertise in photonic-electronic co-design, advanced packaging, and increasingly automated photonic test.

This diversity provides several routes from device innovation to manufacturable systems. Silicon photonics can leverage semiconductor manufacturing infrastructure and scale, while InP enables the integration of active functions such as lasers and amplifiers. SiN provides very low propagation loss for applications where spectral performance or passive functionality are critical.

The opportunity is increasingly about combining the strengths of different materials and integration approaches at system level rather than one platform replacing another. That technological breadth is most valuable when different parts of the ecosystem work together effectively, allowing ideas to move efficiently from research into manufacturable products. Europe already has many of the essential ingredients. But with the US and Asia investing rapidly in AI infrastructure, Europe must act now to expand its manufacturing capacity if it wants to secure a competitive position in this market.

Competitive advantage is changing

As the industry matures, the basis of competition is evolving. For many years, technical performance was the primary differentiator. Strong technology remains necessary, but it is increasingly insufficient on its own. As integrated photonics moves towards larger-scale deployment, competitive advantage is shifting towards the ability to turn technology into qualified, manufacturable products quickly and reliably.

Technical performance will continue to improve and remain important. The differentiator will increasingly be how effectively companies and ecosystems combine process maturity, manufacturing scale, qualification, and supply-chain readiness to move from innovation to deployment.

AI is shaping more than AI infrastructure

Although AI is driving much of today’s investment, I do not expect its impact to stop there. The funding data reflects this. While AI accounts for the majority of strategic investment activity, around 40% of disclosed funding rounds continue to support broader integrated photonics applications.

Healthcare, industrial sensing, mobility, quantum technologies, and defence all represent important opportunities. While these markets will mature at different speeds, advances in manufacturing, packaging and supply chains driven by AI could accelerate their development considerably. European companies need to prepare for that demand now, rather than risk their early advantages.

Responding at the pace of the market

When I talk about urgency, I am not referring to rushed development or shortcuts in engineering. Integrated photonics will always depend on rigorous research, careful validation, and close collaboration across the ecosystem.

Urgency means recognising how quickly the market has changed. Investment is accelerating, supply chains are being reorganised, and companies are making strategic decisions earlier because they understand that future leadership is being established today.

Europe has invested consistently in integrated photonics for decades, creating an ecosystem that spans the entire value chain. But without timely investment in manufacturing capacity, Europe risks seeing its technologies developed into commercial products elsewhere.

Industry and government must act together. Companies need to invest in scaling production, while governments must create the conditions for that investment – including infrastructure, skills, and targeted public funding where private capital falls short.

Europe has the capabilities to lead in integrated photonics. The question is whether we can mobilise them quickly enough to secure our position before others do.