China Securities Co., Ltd. has released a research report indicating that driven by AI, the global PCB industry is entering a new upward cycle. With the demands of orthogonal backplane technology and the Covop process upgrade, the future of PCBs will increasingly resemble semiconductors, with their value steadily rising. Companies like Meta, Amazon, and Google, which have weaker in-house chip design capabilities compared to Nvidia, place higher performance requirements on materials like PCBs, making their value proposition more elastic. As the need for short-distance data transmission continues to grow, PCBs are undergoing continuous upgrades, which in turn drives the evolution of the upstream supply chain. For instance, copper-clad laminates are advancing from M6/M7 to M8/M9 grades. As domestic PCB companies increase their global market share, this is also promoting the localization of the upstream supply chain. Starting from copper-clad laminates, this trend is further boosting the domestic market share of high-end materials such as resins, glass fiber cloth, and copper foil.

The original text is as follows:

AI Drives PCB Demand Upgrades, Accelerating the High-End and Localization of Dry Film Photoresist

Dry film photoresist is a film material that solidifies upon exposure to ultraviolet light, primarily used for circuit transfer in the PCB manufacturing process. The global PCB and packaging substrate industry has entered a structural growth cycle driven by AI servers, data centers, and high-speed network equipment. This corresponds to a compound annual growth rate (CAGR) of 7.7% from 2025 to 2030. Demand for high multi-layer boards, HDI, and packaging substrates is expected to grow faster, with projected CAGRs of 8%, 9.2%, and 10.9% respectively for the same period. In contrast, growth for traditional consumer electronics, standard single/double-sided boards, and general FPCs is expected to be slower, with CAGRs of 2.8% and 3.8% between 2025 and 2030. The future core growth for dry film photoresist will primarily come from high-end applications like servers/data storage and AI PCBs, rather than traditional electronics like phones and home appliances. This shift will increase the average selling price and consumption volume of dry film photoresist. The main reasons are: 1) Increased Volume: Traditional servers typically use 12-26 layer CPBs, while AI servers generally use 22-layer or higher HDI boards and 30-layer or higher multi-layer boards, with a maximum layer count reaching 24-78 layers, and potentially rising to over 80 layers. Each layer of a multi-layer board requires exposure, development, and etching, so the number of pattern transfer steps increases correspondingly, significantly boosting the usage of dry film photoresist. 2) Higher Unit Price: This is driven by increased demands for resolution and yield. In terms of resolution, as line widths and spacings develop from about 4 mils towards 3.5 mils and finer, high-end dry film photoresist requires higher resolution, adhesion, and other properties, significantly raising performance requirements. Regarding yield, because AI PCBs are large in area, have many layers, and are high in value, a defect in a single layer can lead to the scrapping of the entire board, thus increasing the requirements for yield.

Combining the development trends of the PCB industry, the report estimates the total addressable market for dry film photoresist from 2026 to 2030 to be 9.5 billion, 10.4 billion, 11.3 billion, 12.4 billion, and 13.6 billion yuan respectively, corresponding to a CAGR of about 9.4% from 2025 to 2030. The breakdown is as follows: (1) Single/Double-sided boards, FPC: Products are homogenized with intense price competition. Demand is expected to grow modestly with ordinary electronics like phones and home appliances, corresponding to a CAGR of ~3% from 2025 to 2030. (2) Multi-layer boards: AI is driving rapid growth in products with 18+ layers and 30+ layers, increasing the consumption of dry film per unit. The growth rate is expected to be higher than that of multi-layer PCBs, with a CAGR of ~9% from 2025 to 2030. (3) HDI: Driven by the replacement of wet film with inner-layer dry film, the penetration of LDI, and upgrades to high-resolution products, the value of dry film is increasing. This segment's growth is expected to surpass that of multi-layer PCBs, with a CAGR of ~11% from 2025 to 2030. (4) IC Substrates: AI GPUs, ASICs, server CPUs, and advanced packaging are driving the growth of ABF/BT substrates. Simultaneously, large sizes, multiple build-up layers, and fine-line electroplating are increasing the value of dry film. This is expected to be the fastest-growing dry film photoresist category, with a CAGR of ~12% from 2025 to 2030.

Domestic Manufacturers' Market Share Continues to Rise, Multiple Factors Favor Domestic Substitution of Dry Film Photoresist

Since 2023, the sales volume of dry film photoresist products from domestic manufacturers like Chuyuan New Materials and Fuxster has been growing rapidly and consistently. Chuyuan New Materials' dry film sales from 2023 to 2025 were 2.16, 2.70, and 3.48 billion square meters respectively, with year-on-year growth rates of 2.56%, 25.24%, and 28.61%. Fuxster's dry film sales for the same period were 1.15, 1.59, and 1.89 billion square meters, with year-on-year growth rates of 4.52%, 37.97%, and 18.67%. The sales growth rate of dry film for these two manufacturers is significantly higher than the industry average for the same period. Looking ahead, the report expects the trend of domestic substitution for dry film photoresist to continue. This is mainly because leading PCB companies have already started adopting domestic dry film photoresist in volume, and the downstream PCB production capacity is concentrated in China, providing numerous opportunities for the introduction of domestic dry film photoresist.

Risk Warning

1. Risk of lower-than-expected end-user demand. If the PCB industry's prosperity or the growth of downstream demand for AI servers, high-end HDI, and packaging substrates falls short of expectations, the sales volume, product structure upgrades, and profitability improvements of dry film photoresist could be lower than anticipated. The growth in the PCB market is also significantly structurally divergent, with weak demand from traditional consumer electronics and ordinary PCBs, meaning the industry's growth benefits may be concentrated among a few high-end suppliers.

2. Risk of product development and volume ramp-up falling short of expectations. High-end dry film photoresist requires high resolution, adhesion, plating resistance, and process compatibility, leading to long customer qualification cycles. If Fuxster's progress in introducing and scaling up its products within server PCB and substrate customers is not as expected, or if Chuyuan New Materials' R&D achievements in high-end areas like IC substrates are not smoothly commercialized, it could impact revenue growth and gross margin improvement.

3. Risk of intensified market competition. Overseas and Taiwanese market leaders still hold advantages in technology, branding, and customer certification in the high-end market. Accelerated capacity expansion by domestic manufacturers could intensify price competition for general-purpose products, leading to downward pressure on average selling prices and unit profitability. If new capacity fails to secure sufficient orders, it could also lead to lower capacity utilization rates and increased depreciation pressure.

4. Risk of raw material price fluctuations. Fluctuations in raw material prices, slower-than-expected progress in the self-supply of core raw materials, demand fluctuations from major customers, and changes in overseas market expansion and trade policies could all adversely affect the operational performance of the dry film photoresist businesses of these two companies.