Source: Laserfair.com  15^th May 2026

Founded in 2021, Glitterin Tech builds its core competitiveness on silicon photonic chips and integrated optoelectronic technology. In merely four years, it has completed a three-stage leap: from laboratory prototypes to engineering prototypes, and then to mass production of application-ready products. Leveraging the optical convolution spectrometer co-developed with the University of Cambridge, it has redefined the fundamental physical architecture of spectrometers from the bottom up, breaking the industry's long-debated “impossible trinity” of miniaturization, high performance and low cost.

Recently, Glitterin Tech has secured backing from prominent investors including CASSTAR, MiraclePlus, Shaanxi Provincial Government Guidance Fund and YOFC Industrial Investment. Its products cover scenarios ranging from consumer electronics wearables to industrial in-line inspection, with in-depth cooperation established with Xiaomi and leading domestic liquor enterprises. Meanwhile, the company has laid out its strategic presence across three core tracks — optical sensing, optical switching and optical computing — driving industrial commercialization through technological innovation.

In this feature, Laser Manufacture News speaks with Ming Liang, Co-CEO of Glitterin Tech, to delve into the company's growth trajectory, core technological breakthroughs, commercialization layout and future strategy, exploring the breakthrough path of Chinese deep tech enterprises in the field of optical chip integration.

Laser Manufacture News: Glitterin Tech has completed the leap from laboratory research to industrialization in just over four years since its founding. Could you walk us through the key stages of the company's development? Which milestone left the deepest impression on you? And what was the biggest challenge in moving from technology validation to engineering implementation?

Ming Liang, Co-CEO of Glitterin Tech: Thank you very much for the attention from Laser Manufacture News. Glitterin Tech's development can be clearly divided into three phases.

The first phase was the exploration of productization solutions, centered on nailing down the right technology roadmap. Should we adopt active or passive chips? How should the optical link be designed? Every decision shaped the final form of our products.

The second phase was the iteration of engineering prototypes, where we gradually evolved from bulky benchtop prototypes into nail-sized wearable modules.

The third phase is application-oriented product development and mass production ramp-up. Multiple products have now entered pilot production.

The biggest pitfall was not a specific technical problem, but our strategic orientation in the early stage of product miniaturization. Our team comes from a cutting-edge research background, and we initially believed we could turn a discrete lab system into a wearable product in just a few months. It turned out there were far too many uncontrollable variables.

We hit the pause button in time: we first refined and solidified the performance of our benchtop prototypes, then advanced step by step with component co-packaging and power consumption reduction. In the end, we delivered a nail-sized sensor with performance on par with benchtop equipment. Deep tech entrepreneurship must respect the laws of engineering scaling.

As Co-CEO, my core responsibility is not to micromanage every technical detail, but to steer our R&D pace and strategic direction, aligning technological research with market demand and financing milestones.

Laser Manufacture News: The company has earned recognition from top-tier investors including CASSTAR and MiraclePlus. What key strengths of Glitterin Tech do you think appeal most to capital partners? How will this funding accelerate the company's growth?

Ming Liang: What investors value is not just how "sophisticated" our technology is, but that we are a company truly rooted in fundamental technologies with the courage to pursue disruptive innovation.

Our goal is not incremental improvement — it is to reshape the landscape of traditional invasive health monitoring with our self-developed on-chip spectroscopy technology. Apple internally calls innovations of this scale "moonshot projects", and many dismissed ours as a pipe dream. But starting from first principles, we have turned it into reality step by step.

In addition, the shared conviction and strong cohesive spirit across the entire team is another factor that deeply impressed our investors.

All funding is channeled into the dual engine of technology R&D and product commercialization. We run two categories of R&D programs: one is forward-looking technology research for the next 3–5 years to maintain our leading edge; the other is product development targeting current market demand.

We consistently allocate around 30% of our resources to advanced research. This model of "keeping one foot on the gas while looking ahead" allows us to respond nimbly to market needs while continuously building a solid technological moat.

Laser Manufacture News: The optical convolution spectrometer co-developed by your company and the University of Cambridge is hailed as redefining the physical architecture of spectrometers. What exactly does it "disrupt"? And what is the fundamental difference from traditional solutions?

Ming Liang: At the core of this achievement is that we have completely discarded the hardware logic of traditional spectrometers, which relies on physical spectral splitting or interferometric scanning. For the first time, the mathematical principle of the Convolution Theorem serves as the sole physical foundation of a spectrometer. This is not an incremental improvement — it is a fundamental reengineering.

Compared with traditional benchtop spectrometers, our solution represents a shift from a hardware-stacking approach to a software-defined architecture. To achieve high resolution and wide bandwidth, traditional systems require long optical paths and large gratings, resulting in bulky size and high costs. In contrast, our simple multi-stage cascaded interferometer structure enables centimeter-scale packaging and a 500nm wide bandwidth, with a total cost of only around 10 US dollars.

Compared with commercially available computational reconstruction micro-spectrometers, our solution moves from "computationally intensive and distortion-prone" to "linear and highly accurate". Those solutions essentially work like assembling a jigsaw puzzle: they require heavy computation and are prone to signal distortion. Ours, by contrast, is inherently linear based on the Convolution Theorem. The calculation can be completed on a microcontroller unit (MCU) within 50 milliseconds, and its accuracy even surpasses that of many commercial benchtop spectrometers.

The most essential difference is our time-for-space trade-off: no complex optical paths or moving components are required, and resolution can be exponentially improved by increasing the number of cascaded stages, with theoretically unlimited bandwidth.

Laser Manufacture News: What core product lines has the company planned based on this technology? What tangible progress has been made in industrial inspection and consumer electronics?

Ming Liang: We have built a full-chain product portfolio from core components to end-user applications. At the core layer are spectral chips, the heart of all our products. For the consumer electronics layer, we have developed micro spectral sensing modules that are button-sized with mW-level power consumption. Combined with AI algorithms, they can real-time monitor skin moisture, blood alcohol levels, blood lactate levels and even blood glucose trends. For the industrial inspection layer, we offer industrial-grade spectrometers. Our multi-parameter fermented grain analyzer cuts laboratory testing time from hours down to minutes. We also provide all-in-one computing boards at the ecosystem layer to facilitate secondary development for our customers.

In consumer electronics, we have established in-depth cooperation with leading manufacturers including Xiaomi, with smooth progress in module integration testing and algorithm optimization. In the industrial sector, we have partnered with multiple leading domestic liquor enterprises, and validation results have far exceeded expectations. Going forward, the technology will also be widely applied in grain storage, feed production, tobacco and other fields.

Laser Manufacture News: What is the biggest challenge for micro-spectroscopy technology to become a standard feature in smartwatches? How will you overcome the hurdles in system integration, power consumption control and algorithm calibration?

Ming Liang: The biggest challenge is not miniaturizing the device itself, but eliminating complex background noise caused by individual physiological variations within an extremely compact form factor, and accurately extracting faint biomarker features from low signal-to-noise ratio (SNR) signals.

We address this systematically across three dimensions.

In terms of system integration, all complex interferometers are fully integrated onto silicon nitride (SiN) photonic chips to form a complete micro spectral sensing unit.

For power consumption control, our linear algorithm based on the Convolution Theorem completes computation on an MCU in under 50 milliseconds, delivering ultra-low power consumption at the mW level.

For algorithm calibration, the inherent linearity of convolution ensures noise is not amplified, allowing the device to withstand wide temperature fluctuations. Meanwhile, we build tailored AI models for different biomarkers to enable precise tracking.

Laser Manufacture News: Beyond optical sensing, the company has also established a presence in optical computing and optical switching. What roles do these three business lines play respectively? Which direction will be your top priority in the next 2–3 years?

Ming Liang: Optical sensing, optical switching and optical computing are the three core tracks in our strategic layout, and we adopt a phased, prioritized approach to roll them out.

Optical sensing is our current core foundational business. It has entered the mass production ramp-up phase and is about to deliver early returns on investment. Over the next 2–3 years, we will continue to deepen our presence in this field and rapidly expand our market share in consumer electronics and industrial inspection.

Optical switching is the core driver of our second growth curve. Amid the boom of large AI models, rapidly reconfigurable silicon photonic switching solutions have become the key to improving throughput efficiency in intelligent computing centers. We have listed it as another major R&D focus and are conducting technical validation with clients.

Optical computing serves as our long-term technology reserve. Leveraging the team at the University of Cambridge, we are developing optical ASIC chips and will quickly enter the market once the industrial chain matures.

In the next 2–3 years, our top priority will be scaling up mass production of optical sensing, while accelerating the commercial rollout of optical switching.

Laser Manufacture News: What is Glitterin Tech's strategic layout in telecom optical switching and AI optical acceleration algorithms? Will this become your second growth curve following spectrometers?

Ming Liang: Our founding team has deep roots in integrated optoelectronics spanning many years. Our founder Professor Cheng Qixiang boasts 16 years of research experience in optical switching, and previously led his team to develop the world's first 16×16 optical switch chip. This accumulated expertise forms the core foundation for our rapid entry into the AI photonics track.

In telecom optical switching, we have brought mature technologies to market and are conducting joint validation with multiple leading intelligent computing center clients. As the critical "gateway" for information throughput, optical switching will undoubtedly become Glitterin Tech’s most important second growth curve after spectrometers.

For AI optical acceleration algorithms, we are developing optical ASIC chips leveraging our Cambridge-based team, following a strategy of proactive R&D positioning and market readiness. once the industrial ecosystem matures, we will be able to roll out commercial applications rapidly.

Laser Manufacture News: What technological iterations will the company roll out in hardware, algorithms and mass production over the next 3–5 years?

Ming Liang: Over the next 3–5 years, we will drive iterations across three layers, centered on our two core directions of optical sensing and optical switching.

For optical sensing hardware, we will advance the on-chip integration of optical components, aiming to reduce power consumption to below 1mW, and accelerate high-value applications such as non-invasive blood glucose monitoring and early cancer screening from the chip level. On the algorithm front, we will build a larger-scale human physiological database to address the generalization challenge across different individuals. For mass production, we will optimize chip design to improve yield, and build a proprietary Class 10,000 clean packaging and testing line.

On the optical switching front, we will accelerate the commercialization of silicon photonic optical circuit switching (OCS), and continue pre-research on 3D-printed hybrid integration and AI optical computing to build reserves for our third and fourth growth curves.

Laser Manufacture News: What is the capacity plan for your Class 10,000 clean packaging and testing workshop? Under the target of exceeding 100 million RMB in revenue within two years, what is your market expansion strategy?

Ming Liang: The Class 10,000 clean packaging and testing workshop we are building is expected to meet the production demand for 100,000 sets of sensing modules over the next two years. More importantly, building the production line is also a process of establishing mass production process standards. once the process is fully validated, mature workflows can be rapidly replicated whether through in-house expansion or cooperative production.

Our market expansion follows a three-pronged strategy: focus on leading clients, build an open ecosystem, advance domestic substitution. We will concentrate resources on serving top-tier clients and build reputation through benchmark cases; open up our spectral modules and algorithm platform to co-develop application scenarios with upstream and downstream partners; and prioritize cooperation with domestic industrial chains to promote the localization of the silicon photonic industry chain.

Laser Manufacture News: How do you view the gap between China and the global state-of-the-art in high-end optical chips? What is the strategic significance of Glitterin Tech’s breakthroughs for domestic substitution?

Ming Liang: To put it objectively, China has no notable gap with global advanced levels in chip design capability. The disparity mainly lies in the maturity and stability of chip manufacturing processes. Many domestic companies still source wafer fabrication from international foundries — not because domestic production is technically unfeasible, but because they cannot afford the time risk of process failures.

The good news is that silicon photonic chips do not require cutting-edge process nodes. 60nm and 90nm processes are fully sufficient, and there is no generational gap in manufacturing equipment. Several domestic fabs equipped with 45nm lithography tools are already developing 90nm silicon photonics processes. Since our products entered mass production, our annual wafer throughput has exceeded 1,000 wafers, and we now primarily use domestic foundries for fabrication.

As commercialization advances, we will bring more mass production volume to domestic fabs, accelerating the maturity of domestic silicon photonics processes and encouraging more companies to shift their fabrication onshore. This is not just a success for Glitterin Tech alone, but a powerful driver for the localization of China’s entire silicon photonics industry chain.

Laser Manufacture News: Looking ahead five years, how will spectroscopy technology transform everyday life? What insights can you share with deep tech entrepreneurs?

Ming Liang: Over the next five years, spectroscopy technology will evolve from "precision laboratory instruments" into "ubiquitous embedded sensors". It will become a personal "material identification expert" for everyone: verifying whether coffee is made from genuine Arabica beans, or identifying plastic types at the touch of a button for waste sorting.

More importantly, it will turn smartwatches into portable personal clinics, monitoring indicators such as blood glucose, blood lipids and uric acid in real time, truly delivering on the vision of proactive preventive healthcare.

For deep tech entrepreneurs, I would like to share two key lessons. First, return to first principles to build the foundation of disruptive technology. Instead of making incremental improvements within existing frameworks, innovate from the underlying logic — only then can you build a genuine technological moat.

Second, pursue extreme system-level integration and cost advantages. Advanced technology is only the first step; real-world deployment and scalability are what define success. There are no shortcuts in deep tech entrepreneurship. But as long as you stay true to your founding mission, and develop truly valuable technologies and products rooted in first principles, you will surely create landmark achievements for China’s deep tech sector.