In-Depth Analysis of the Silicon Photonics Industry: From “Light in the Laboratory” to “Product in the Factory” – What’s in the Middle?
Hello everyone, I’m your financial journalist. Today, we’re going to talk about a topic that might sound a bit technical, but it’s actually quietly transforming the speed of the internet and even the computing power of AI: the silicon photonics industry.
If you follow tech news, you might have heard terms like “optical communication” or “CPO (Co-Packaged Optics). Simply put, current internet data transmission relies on electrical signals traveling through copper wires. However, as speeds increase, copper wires become increasingly inefficient and generate a lot of heat. That’s where the idea of using “light” for data transmission comes in – similar to how fiber optics are used in homes – by integrating optical components directly into chips.
Recently, at the 2026 Pujiang Innovation Forum, industry leaders discussed a very practical issue rather than boasting about the advanced technology: how to actually produce, sell, and use these products?
It’s like everyone can cook a new dish (the technology is feasible), but the real challenge is how to open a chain of restaurants that thousands of people can afford and enjoy. That’s where the real difficulty lies.
Below, I’ll break down this news into five key points to explain in plain language where the silicon photonics industry is stuck and how it can overcome these obstacles.
---
1. Manufacturing Challenges: Not the Ability to Produce, but the Quantity and Stability of Production
Core Issue: Domestic manufacturing platforms are still developing, and startups need to work alongside them to improve their capabilities.
The news highlights a stark reality: China still lags behind international leaders in silicon photonics manufacturing and advanced packaging.
It’s like wanting to open a new restaurant, but the local central kitchen (the manufacturing platform) is still under construction, and the equipment isn’t fully operational. For startups, you can’t just show the design drawings and expect the factory to produce one million units immediately, as the production process and quality rates are not yet stable.
Nie Hui, the general manager of Imec, put it bluntly: The biggest problem right now is mass production. This means startups can’t just act as clients; they need to collaborate with manufacturing platforms to refine the processes. This increases costs due to trial and error, but it also presents opportunities – the first company to succeed will set the standards.
Plain Language: In the past, factories set the rules, and you followed them. With silicon photonics, both the factory and the startup need to work together to establish these rules. It’s like a dance where both parties need to coordinate well; otherwise, mistakes can occur. That’s why places like Shanghai are building “shared cleanrooms” and testing platforms to help startups identify issues before large-scale production.
---
2. The Role of Pilot Production Lines: Bridging the Gap from “0 to 1” to “1 to 100”
Core Issue: Lab samples work perfectly, but mass production fails. There’s a lack of a “translator” to bridge this gap.
The news mentions the importance of pilot production lines (Pilot Lines). Researchers often create perfect chips in the lab, but when scaled up, minor variations in temperature, pressure, or material purity can ruin the products. Pilot lines bridge this gap by providing a testing environment.
Xia Xiaoyuan from the Shanghai Institute of Microtechnology says pilot lines are essential for engineering development – they’re not as precise as labs but not as rough as factories.
Plain Language: Think of a driving school: Labs teach you the theory, factories require high-speed, stable production, and pilot lines are the practice area where you can make mistakes and refine your skills.
Without this step, startups might fail when transitioning from lab to factory production. Shanghai’s pilot lines are working on 11 projects to help transform designs into viable products.
---
3. Fragmented Industry Chain: Lack of Unified Standards
Core Issue: Different parts of the chain (materials, manufacturing, packaging) don’t communicate effectively, leading to inefficiencies and high costs.
The news emphasizes the lack of standardization. The silicon photonics industry is like a fragmented market with various players working independently. For example, materials and packaging may not be compatible with chip designs.
Xia Xiaoyuan suggests starting from the end-users to standardize interfaces and requirements. If different phone brands use different charging ports with varying power standards, users would face inconvenience. The same issue applies to the silicon photonics industry.
Solution: A leading role is needed to establish common standards for dimensions, connectors, and other specifications, allowing for smoother integration of components.
---
4. Focusing on Market Demand: Technology Is Important, but So Is Who Will Use It
Core Issue: Great technology needs a large market demand for mass production to be viable.
Cai Jinxu from the Shanghai Future Start-up Fund pointed out that a specific application scenario is crucial. If there’s no clear market demand, even advanced technology won’t be successful. For example, AI data centers need high-bandwidth, low-power chips, while consumer electronics require different specifications.
Plain Language: You can’t just develop a high-performance engine and then ask who will buy it. You need to understand the market’s needs – whether it’s for long-distance SUVs or compact cars. The industry should start with the end-users’ requirements and design accordingly.
---
5. The New Role of Capital: More Than Just Financing, but as a Connector of an Ecosystem
Core Issue: The silicon photonics industry has a long cycle and requires significant investment. Single companies can’t handle it all; capital acts as a catalyst.
The news mentions the Shanghai Future Industry Fund’s investment of over 900 million yuan in the industry, forming an alliance of eight funds to collaborate. This shift indicates that capital is no longer just about providing funding; it’s about coordinating resources and supporting the entire ecosystem.
Plain Language: Silicon photonics is too complex for a single company to handle all aspects. Funds help select promising teams, bring together necessary resources, and share risks, thus accelerating the industry’s development.
---
Conclusion: The “Coming-of-Age” of the Silicon Photonics Industry
In summary, this news signals that the silicon photonics industry has moved beyond the stage of showcasing technology to focusing on practical implementation:
- In the past, the focus was on who published the most research papers or had the best lab samples.
- Now, the focus is on the stability of manufacturing platforms, the speed of pilot lines, the standardization of processes, the accuracy of application scenarios, and the strength of the capital ecosystem.
For consumers, this means faster internet speeds and more efficient AI systems. For investors and industry players, the era of working alone is over; collaboration is essential for the industry to scale commercially.
Shanghai’s success in the silicon photonics industry isn’t due to its number of leading companies, but rather because it’s building a comprehensive infrastructure – from cleanrooms and pilot lines to standardization and capital alliances. This is the true foundation for the industry’s growth.
The silicon photonics industry is entering a new phase where practical solutions and market-driven approaches are key to its success.