Why silicon photonics will help power the UK’s AI hardware future

Why the AI Hardware Plan matters 

The Plan recognises that sovereign capability in chips and semiconductor technologies is strategically important. As AI systems demand more, countries that can scale enabling hardware will be better placed to secure resilience and long-term competitiveness. 

That is especially true in the UK, where there is already capability across the photonics stack. We have world-class research, a growing innovation ecosystem and a strong base of companies working in semiconductors, optics and advanced computing. 

Dr Andy Sellars, Director of Strategy at CORNERSTONE and Chair of the Semiconductor Expert Working Group, says the Plan represents a critical step towards securing Britain’s future capability in the technologies that underpin AI. 

“The UK must strengthen its sovereign AI capabilities and work with allies, or risk losing out in the global race. In prioritising AI hardware development, the government has rightly recognised the strategic importance of chip development to support the country’s AI ambitions. It’s also encouraging to see increasing recognition from governments and industry alike of silicon photonics’ ability to mitigate AI power consumption.” 

Dr Andy Sellars, Director of Strategy at CORNERSTONE and Chair of the Semiconductor Expert Working Group

Reducing the growing energy demands of AI systems 

Recent investment activity across these industries has reinforced the idea that photonics could play a central role in next-generation computing infrastructure.

Today’s AI systems are limited by how quickly and efficiently data can move between processors, memory and data centres. Optical connections built using silicon photonics can transmit data far more efficiently than traditional electrical connections, dramatically reducing latency and power consumption. 

Photonics uses light rather than electrons to transmit and process information, offering advantages in bandwidth, density, speed and energy efficiency. That is why silicon photonics represent one of the most significant opportunities for industrial growth this decade, and a key enabling technology in critical hardware infrastructure. 

This is becoming more important as AI workloads grow. Data centres already consume vast amounts of energy, and as AI adoption accelerates, that demand is projected to use twice as much energy by 2030. 

The AI applications that silicon photonics is transforming 

Silicon photonics is already playing a crucial role in cloud AI, particularly in routing data within data centres. But its greatest opportunity may lie in edge AI, where lower latency and power consumption are essential.  

Dr Andy Sellars says, “Photonic integrated circuits can deliver optical compute functions with comparatively low energy consumption, making them a strong fit for emerging AI hardware needs. 

This, alongside emerging areas such as neuromorphic computing – which mimics the structure of the brain – and optical processing, is where the UK has strong capability across the photonics stack.” 

The opportunity shines bright. According to our research, 56% of UK companies developing or planning to develop silicon photonics chips are looking at AI hardware applications, revealing just how closely the UK’s silicon photonics ecosystem is aligned with the demands of future AI infrastructure. 

Where the UK can lead 

The UK does not need to replicate every part of the global AI hardware supply chain to be competitive. It does, however, need to focus on the areas where it has genuine strengths and where targeted investment can make the difference. 

We believe the UK has a clear opportunity to build on its strengths and secure a globally competitive position in next-generation AI infrastructure. But to do that, it needs the right foundations: sustained investment, stronger pathways from research to scale-up, and infrastructure that helps promising technologies move faster. 

That’s why we’re making the case for a national silicon photonics pilot line. A dedicated pilot line would give UK companies better access to the facilities they need to prototype, integrate and scale advanced photonic devices. It would strengthen the wider ecosystem and help ensure that more of the value created in the UK is captured here. 

The UK’s AI Hardware Plan is an important signal of intent. Now the challenge is to build on that momentum and turn it into capability. 

Silicon photonics has a central role to play in this journey. It can help address the power and performance limits of current AI hardware, support infrastructure and strengthen the UK’s position in a globally competitive market. With the right investment and infrastructure, the UK can move from strong research to real leadership in the applications that matter most. 

To learn more about CORNERSTONE’s pilot line vision, read more here.