Morgan Stanley released its latest Greater China tech hardware sector report on September 17, stating that while co-packaged optics (CPO) could potentially reduce the usage demand for premium copper clad laminate (CCL) over the long term, this technology is unlikely to deliver a substantial blow to the premium CCL market in the near term. Given that CPO still faces numerous challenges in areas such as cost, manufacturing yield, reliability, maintainability, and thermal management, the investment bank projects that the impact on premium CCL demand between 2026 and 2028 will be limited. The robust demand for high-performance CCL driven by AI systems, along with the ongoing tight supply landscape, is expected to persist.
On September 17, shares of Taiwan-based premium CCL manufacturers Taiwan Union Technology Corp and Elite Material Co Ltd declined 7.7% and 6.7%, respectively, while the broader Taiwan Weighted Index gained 0.8% during the same session. Following these notable pullbacks, market attention has refocused on a pivotal question: as CPO technology accelerates its adoption into AI infrastructure, could demand for premium CCL be meaningfully undermined?
Addressing CPO Concerns: Shorter High-Speed Signal Paths and the CCL Demand Question
The concern primarily stems from the architectural shifts inherent in CPO. In traditional AI server and switch architectures, high-speed electrical signals must travel through the printed circuit board (PCB) over relatively long distances between the chip and the optical module. As data transmission rates continue to climb, reducing signal loss requires PCBs to incorporate premium CCL materials with ultra-low dielectric constant (Dk) and ultra-low dissipation factor (Df). CPO's core concept, however, involves integrating the optical engine and the ASIC chip more closely near the packaging substrate, thereby significantly shortening the high-speed electrical signal pathways that would otherwise traverse the PCB. In theory, this could lessen the system's reliance on ultra-low Dk/Df high-speed CCL, raising fears that rising CPO adoption might erode the per-unit value of premium CCL in AI systems.
Morgan Stanley, however, emphasizes that this is not a new industry debate. The bank has previously examined the potential long-term effects of CPO on high-speed CCL demand and maintains that CPO represents a legitimate long-term risk for the premium CCL sector, but it is not an imminent threat to demand.
Multiple Hurdles Hinder Large-Scale CPO Adoption
Morgan Stanley believes the market may currently be underestimating the engineering challenges required for CPO to achieve large-scale commercialization, particularly within scale-up architectures in AI clusters. The report points out that broader CPO adoption still needs to resolve issues related to cost, manufacturing yield, reliability, maintainability, and thermal management. Consequently, even though CPO holds the potential to diminish premium CCL usage over the long term, the risk of it broadly weakening overall CCL demand for AI systems before 2028 remains limited. This suggests that within the roughly two-year outlook, the rapid expansion of AI computing infrastructure will continue to serve as the primary driver of premium CCL demand, and the potential material substitution effect from CPO is not yet sufficient to alter this trajectory.
Morgan Stanley therefore retains a constructive view on premium CCL fundamentals. The report explicitly states that it currently sees no clear impact from CPO on the near-term CCL demand or earnings forecasts of Elite Material Co Ltd and Taiwan Union Technology Corp, and its earlier assessment of persistently tight supply conditions for premium CCL remains unchanged.
CPO as a Long-Term Risk vs. Near-Term Supply-Demand Reality
Over a longer time horizon, Morgan Stanley does not dismiss the structural influence CPO could exert on the premium CCL industry. As CPO technology gradually matures and optical engines move closer to ASICs, the distance high-speed electrical signals must travel through PCBs will decrease, potentially affecting both the unit volume and the specification upgrade demand for premium CCL. The bank acknowledges, therefore, that broader CPO adoption represents a credible long-term risk to the value proposition of premium CCL.
However, the report draws a clear distinction between "long-term technology risk" and "demand trends over the next two years": within the 2026 to 2028 forecast window, the pace of CPO penetration is not sufficient to significantly alter the demand outlook for premium CCL. In other words, the long-term technology substitution logic currently being traded by the market has not yet translated into actual short-term order or earnings pressure.
It is worth noting that Morgan Stanley retains a neutral stance on the Greater China tech hardware sector overall. Meanwhile, among the companies covered in the report, as of September 16, the bank maintains an "Overweight" rating on both Elite Material Co Ltd and Taiwan Union Technology Corp.