Not Every Light Chip Becomes a Breakthrough
Martijn Heck on why photonics will shape the future of chips—but not replace electronics anytime soon
In a sector driven by bold claims around AI and next-generation chips, Professor Martijn Heck offers a more grounded perspective: photonics matters — but without scale, reliability and clear demand, it remains a niche within an increasingly complex semiconductor landscape.
Before discussing chips, Martijn Heck raised a different concern: the growing difficulty of trusting what we see and read in technology discourse. AI-generated images, unchecked claims and recycled narratives are eroding credibility—even within scientific communities.
That skepticism carries into his view of the semiconductor industry. Where many see disruption and breakthrough, Heck sees a field shaped by constraints: manufacturing limits, reliability requirements and market demand.
Photonics, in that sense, is not a revolution waiting to replace electronics—but a technology finding its place within a broader system.
“The future of chips is not purely electronic and not purely photonic, but hybrid.”
Martijn Heck, Professor of Photonic Integration, Eindhoven University of Technology
Electrons Compute. Photons Carry.
At its core, the distinction is straightforward.
Electronic chips, based on the movement of electrons, remain unmatched in computation: logic, processing and control. Photonic chips, using light (photons), excel in transporting information—offering high bandwidth and potentially lower energy consumption.
This division of labor is key.
Rather than replacing electronics, photonics complements it. The emerging architecture is therefore hybrid: electrons handle computation, photons handle communication.
From Scaling Down to Combining Technologies
For decades, the semiconductor industry was defined by one question: how to make transistors smaller.
That question still matters—but it is no longer sufficient.
Today, the industry is evolving along multiple dimensions:
- Heterogeneous integration (combining different chip types)
- Chip stacking and packaging
- Material diversification (silicon, InP, GaAs)
- Improved interconnects between chips
The result is a more layered and complex market, where innovation is no longer confined to a single trajectory. Even giants like Intel are now looking at imec and partners in Europe to bridge the gap between CMOS and photonics.
As Heck suggests, the real shift is not just technological—it is architectural.
Photonics: Promising, but Still Niche
Despite its potential, photonics remains largely a niche technology.
Applications exist across:
- data communication and fiber networks
- sensing and biosensors
- defense systems
- medical imaging (e.g. OCT)
- LiDAR and industrial scanning
But none of these, yet, define a broad, dominant volume market.
“Photonics is interesting, but without volume and reliability it remains a niche.”
Martijn Heck
This reflects a fundamental industrial challenge.
Some companies demand large volumes, but production capacity is still limited. At the same time, in several segments only a handful of customers drive demand. That imbalance creates a familiar bottleneck: too advanced for the lab, not yet mature enough for mass deployment.
In semiconductors, technical performance alone is never enough. Reliability, scalability and cost ultimately determine success.
The Market Is Multidimensional
One of the more important insights from the conversation is that the chip industry is no longer a single race.
It is a system of interdependent layers:
- compute (processors, AI acceleration)
- communication (data movement, interconnects)
- sensing (optical and environmental detection)
- specialized applications (defense, healthcare, infrastructure)
Photonics plays a role in several of these layers—but rarely dominates them.
This makes the technology strategically relevant, but not universally transformative—at least not yet.
Europe: Strong in Knowledge, Fragmented in Execution
The European position adds another layer of complexity.
There is no shortage of expertise. Research institutions, specialized companies and regional ecosystems—such as Brainport Eindhoven—form a strong knowledge base.
But compared to the United States or parts of Asia, Europe faces structural challenges:
- fragmented markets
- slower coordination
- limited large-scale industrial consolidation
“Europe has a lot of knowledge, but ultimately depends on collaboration.”
Martijn Heck
This reliance on collaboration is both a strength and a constraint. It enables cross-border innovation, but can slow down decision-making and scaling—especially in a capital-intensive industry like semiconductors.
Initiatives such as chip competence centers aim to bridge gaps between design, testing and production. Their role is not just technical, but economic: lowering barriers for startups and making the ecosystem more accessible to investors.
A Hybrid Future, Not a Replacement Story
The key takeaway from Heck’s perspective is not about technological dominance, but coexistence.
The semiconductor industry is not moving toward a single winning technology. Instead, it is becoming a system in which:
- electronics remain the foundation of computation
- photonics enhances communication and specific functions
- hybrid integration connects both worlds
This shift requires a different way of thinking.
Not in terms of replacement—but in terms of combination.
What Ultimately Matters
In a field often driven by expectations, Heck’s position is notably grounded.
Technologies do not succeed because they are elegant, fast or conceptually superior. They succeed when they can be:
- produced at scale
- trusted in real-world conditions
- integrated into existing systems
- supported by sustained demand
Or, as his perspective throughout the conversation suggests: “In the end, the semiconductor industry doesn’t reward what is possible—it rewards what works at scale.”
Photo credit: Angeline Swinkels / TU/e
