Recently, I noticed an interesting phenomenon—silicon photonics and CPO (Chiplet Packaging and Optics) have suddenly become market focal points, with related stocks being hotly contested from Taiwan stocks to U.S. stocks. Rather than saying this is short-term speculation, it’s more accurate to see it as a wave of genuine industry upgrading.

First, let’s discuss why this sector is so hot. After the explosion of AI computing demands, traditional methods of transmitting signals via copper wires are nearing their limits—too hot, too slow, and too energy-consuming. The emergence of silicon photonics and CPO aims to solve this pain point, with a simple core logic: using light instead of electricity to transmit data. It sounds a bit sci-fi, but the technology has already moved from laboratory experiments to commercialization.

The concept of silicon photonics is to miniaturize large optical components to chip scale, directly integrating them onto silicon substrates. CPO takes this further by no longer making optical modules as “USB drives” plugged into switches, but instead directly packaging them next to CPUs or GPUs, which can save over 30% of energy consumption and significantly increase speed. Why are these two technologies always linked? Because to realize CPO, optical components must become very small and flat, and silicon photonics technology is perfectly capable of achieving this.

Taiwan’s advantage in this wave is quite evident. From wafer foundries to packaging, testing, and optical communication components, the entire supply chain’s completeness is unmatched globally. TSMC is not just a foundry but also defining CPO packaging standards; their COUPLE platform has become an industry core. Entering this year, silicon photonics technology has officially entered mass production, meaning those companies with early deployment and solid technology are beginning to reap benefits.

In Taiwan stocks, TSMC is undoubtedly the leader, as their packaging standards directly determine the entire ecosystem. Advanced packaging and testing companies like ASE and KY, as well as optical communication component manufacturers like Innolux, are key to commercializing CPO. Shunzen, due to its deep involvement in fiber optic connection technology development, benefits the most—this “interface” may seem simple, but precisely guiding light into chips involves extremely high technical barriers. Companies like Boro, Lianya, and Fan Quan, though smaller in scale, have mastered core technologies in their respective niches and are worth watching as silicon photonics low-priced stocks.

The logic in the U.S. stock market is somewhat different. Giants like Broadcom, Marvell, and NVIDIA control core silicon photonics designs and communication protocols, with Taiwanese manufacturers mainly doing foundry work based on their designs. Broadcom’s leading position in CPO is unshakable; their Tomahawk series has become the standard for AI data centers. Marvell competes with Broadcom in optical transceiver chips and recently received billions of dollars in investment and cooperation from NVIDIA. Credo recently acquired DustPhotonics for $1.3 billion, gaining control of photonic integrated circuit technology—this kind of transformation offers far more imagination than just chip design companies. Coherent and Lumentum, traditional leaders in optical transceiver modules, are also accelerating their shift toward silicon photonics solutions.

But a few risks need to be noted. First is yield issues—this is currently the industry’s biggest pain point. CPO packages optical components and chips together; if any part fails, the entire expensive chip may need to be scrapped. Therefore, when reviewing financial reports, pay close attention to gross margin trends—if revenue increases but gross profit declines, it may indicate yield struggles. Second is the risk of specification wars—LPO (Linear Pluggable Optical Modules), as a cheaper and easier-to-maintain alternative, might take market share from CPO before 1.6T becomes widespread. Third, be cautious of companies claiming to be silicon photonics concept stocks but with very low optical communication revenue—these might just be riding the hype. Lastly, don’t forget geopolitical factors—U.S. broadband infrastructure plans will directly impact optical communication demand, and U.S.-China tech tensions could introduce variables.

Recently, many silicon photonics concept stocks have already surged too much. The most reliable way to continue chasing is to watch the earnings calls of major players—once giants like NVIDIA and Broadcom start placing orders with specific Taiwanese packaging or connector companies, that’s a real profit support, not just hype.

Overall, silicon photonics and CPO are not short-term themes but structural growth trends over the next 5 to 10 years. This year marks the critical transition from validation to large-scale mass production, truly testing each company’s technological implementation. The investment logic is clear: in the U.S., watch who sets the standards; in Taiwan, watch who secures orders. While capital chases new themes, returning to fundamentals remains key—prioritize companies certified by major manufacturers with a significant increase in optical communication revenue, so you can seize genuine investment opportunities in this high-speed sector.