Citigroup raised its forecast for the Chinese AI-PCB market, saying that Google's TPU demand may surpass Nvidia's in 2028.

According to Mars Finance, on June 23, Citigroup stated that AI servers and high-speed network upgrades are reshaping the growth prospects of China's PCB industry and raised its target prices for Shanghai Electric Group and Shenghong Technology, citing the potential for continued rapid expansion in AI-PCB demand over the next three years. In a report released on June 21, Citigroup revised its total potential market size forecast for AI-PCB to RMB 152 billion, RMB 307 billion, and RMB 562 billion for 2026, 2027, and 2028, respectively, representing year-on-year growth of 86%, 102%, and 83%. The bank stated that the new model incorporates CPU and optical module PCB demand and provides a 2028 forecast for the first time. Citigroup expects that in 2027, AI-PCB demand will be dominated by ASICs, accounting for approximately 34%; NVIDIA GPU-related demand will account for 24%, CPUs for 16%, switches for 14%, and optical modules for 12%. In terms of growth rate, network-related demand is the strongest, especially with the increased volume of 1.6T optical modules and switches, which may drive optical module PCB demand to grow by 135% and 178% in 2026 and 2027, respectively. The report states that Google's TPU-related PCB procurement may become a key growth driver in 2028. Citi predicts that Google's TPU PCB procurement may reach $16 billion by 2028, exceeding Nvidia's GPU-related PCB demand. Breaking down the new demand from 2027 to 2028, Google TPUs will contribute approximately 30%, Nvidia 25%, and switches and optical modules 14% each. With the rapid expansion of AI-PCB demand, capacity may become a new constraint. Citi predicts that the next round of PCB capacity expansion announcements may begin in the second half of 2026 to support demand in 2028. The bank stated that it typically takes 13 to 15 months for PCB manufacturers in mainland China to go from greenfield projects to mass production, with upstream fiberglass and copper-clad laminate expansion cycles being even longer, potentially reaching 18 months. This means that the supply of AI-CCL may tighten further. Citi believes that with major AI-PCB projects starting in the second half of 2026, the profitability visibility of leading PCB manufacturers with raw material lock-in capabilities will improve in 2027. Related expansion announcements may also act as a catalyst for the industry, strengthening investors' confidence in the long-term growth of AI-PCB. Regarding individual stocks, Citi maintains a "buy" rating on both Shanghai Electric Group and Shenghong Technology. The bank raised its target price for Shanghai Electric Group from RMB 119 to RMB 189, citing the company's strong mass production and delivery capabilities in the fields of data center switches and AI server PCBs. Citi expects Shanghai Electric Group's net profit to be RMB 6.7 billion, RMB 12.1 billion, and RMB 23.2 billion in 2026, 2027, and 2028, respectively, and believes its compound annual growth rate of profit can reach 86%. Citigroup raised its target price for Shenghong Technology from 415 yuan to 456 yuan. Citigroup lowered its 2026 net profit forecast, mainly reflecting lower-than-expected revenue contribution from Rubin, but believes 2027 earnings will remain relatively stable and introduced a 2028 net profit forecast of 23.2 billion yuan for the first time. The bank stated that Shenghong Technology's long-term relationship with Nvidia, HDI capacity, and potential opportunities in data center switches and ASICs still support its valuation premium. However, Citigroup also cautioned that the AI-PCB market still faces multiple risks, including lower-than-expected GenAI-related order share, yield issues, price competition in the automotive supply chain, reduced capital expenditures by cloud vendors, weakening macroeconomic demand, rising material costs, and geopolitical risks between China and the US. Overall, Citigroup's assessment is that AI-PCB demand is no longer solely driven by Nvidia GPUs, but is expanding to ASICs, Google TPUs, high-speed switches, and optical modules. With demand entering the capacity planning phase ahead of schedule in 2028, leading Chinese PCB manufacturers with advanced manufacturing processes, customer resources, and material supply capabilities are likely to continue to be key beneficiaries of the expansion of AI infrastructure.