Pushing the Boundaries of Photonics for Next-Gen AI
Chee Wei LEE
CTO
Elite Photonics Technologies Co. Ltd.
Chee Wei LEE has built a distinguished career at the intersection of scientific research and industrial innovation. His journey in photonics began in 2002 when he was the sole university candidate selected for an internship at Agilent Technologies’ fibre R&D division. That experience sparked a profound interest in optical technology, shaping the trajectory of his career. With a PhD in photonics, he took on a pivotal role at Singapore’s A*STAR, leading research in heterogeneous photonics integration. His commitment to technological advancement led him to the University of Cambridge’s Cavendish Laboratory, where he engaged in pioneering research. However, he recognised the need to bridge scientific discovery with practical applications, prompting his transition into industry. Lee gained hands-on experience in micro-LED displays at a Shanghai-based startup before joining Huawei Technologies, where he focused on silicon photonics integration. These experiences reinforced his ability to translate cutting-edge research into scalable, real-world solutions. Now, as CTO of Elite Photonics, he is at the forefront of developing high-speed photonic chips for data centre transceivers, driving advancements in optical communication. His work is shaping the next generation of data transmission technology, ensuring faster and more efficient connectivity. From research labs to leading industry transformation, Lee continues to push the boundaries of photonics, creating solutions that meet the demands of an increasingly connected world. Speaking with TradeFlock, he discusses his journey and work.
How will photonics transform AI data centres?
AI data centres are reaching a tipping point where traditional electronic interconnects are struggling to keep up. As AI models grow in complexity, the sheer volume of data movement is pushing copper-based interconnects to their limits in both bandwidth and energy efficiency. Photonics is the key to overcoming these challenges. Optical interconnects support terabit-persecond data rates with significantly lower power consumption. Innovations like copackaged optics (CPO) and silicon photonics integrate optical components directly onto chips, eliminating bandwidth bottlenecks and enhancing overall performance. Energy efficiency is another critical advantage. AI workloads demand immense power, much of which is lost in heat dissipation. Photonic interconnects drastically reduce power consumption and cooling needs, making data centres more sustainable. Additionally, optical links enable ultra-fast, long-distance data transfer, reducing latency in AI model training and accelerating workflows. Beyond interconnects, photonic computing is emerging, with optical neural networks and photonic tensor cores promising faster, energy-efficient AI processing. As industry leaders invest in photonics, it will become the foundation of next-generation AI data centres— faster, greener, and more scalable than ever.
What’s the biggest leadership lesson you’ve learnt from managing global tech teams?
Adaptability is the key to leading diverse tech teams. Working across Singapore, the UK, and China, I’ve learnt that leadership isn’t one-sizefits-all—each culture approaches collaboration, decision-making, and innovation differently. Initially, I applied a uniform leadership style, but real progress came when I adjusted to each team’s unique dynamics. Clear communication is equally crucial. Language and cultural differences can create misunderstandings, so I focus on fostering open dialogue and ensuring alignment. Ultimately, trust fuels innovation. When people feel valued and respected, they contribute bold ideas—driving breakthroughs that wouldn’t happen in a rigid, one-dimensional environment.
What key challenges do you face as CTO, and how are you overcoming them?
Bridging the gap between innovation and largescale commercialisation is my biggest challenge as CTO of Elite Photonics. We develop high-speed photonic chips for AI-driven data centres, where demand for bandwidth and efficiency is surging. However, turning cutting-edge technology into scalable solutions presents key hurdles. Scalability is crucial. Silicon photonics holds immense potential, but ensuring highyield manufacturing requires precision in process control, packaging, and supply chain integration. I work closely with fabrication partners to optimise production without compromising performance.Keeping pace with AI’s rapid evolution is another challenge. By collaborating with customers, we align our roadmap with future demands— pushing speed, efficiency, and scalability. Talent acquisition is equally vital. Expertise in photonics and semiconductors is rare, so I focus on building a multidisciplinary team and fostering cross-domain collaboration. Through strategic partnerships, customer engagement, and talent development, we are bridging the gap between research and realworld impact—positioning Elite Photonics at the forefront of next-generation optical connectivity.
What emerging technology excites you most, and how are you leveraging it at Elite Photonics?
The rapid rise of AI-driven workloads is pushing data centre infrastructure to its limits, making co-packaged optics (CPO) one of the most exciting and transformative technology trends today. Traditional copper interconnects can no longer keep up with the bandwidth and energy demands of next-gen computing. CPO, by integrating optical components directly with switch ASICs, eliminates these bottlenecks—delivering unprecedented speed, efficiency, and scalability. At Elite Photonics, we are at the forefront of this revolution, developing high-speed photonic chips and optical engines designed for seamless integration into future CPO architectures. By optimising silicon photonics, enhancing packaging techniques, and collaborating with foundries, we’re not just following industry trends—we’re shaping them. The future of AI and cloud computing depends on breakthrough photonic innovations, and we’re determined to lead the way.
