Citigroup raises China's AI-PCB market outlook, stating that Google's TPU demand may surpass NVIDIA's in 2028

Mars Finance News, June 23 — Citibank said that AI server and high-speed network upgrades are reshaping the growth outlook for China’s PCB industry. It has raised the target prices for Huatian Technology and Shenghong Technology, saying that AI-PCB demand may continue to expand rapidly over the next three years.

In a report released on June 21, Citibank raised its forecasts for the total available market size of AI-PCBs from 2026 to 2028 to 152 billion yuan, 307 billion yuan, and 562 billion yuan, respectively, corresponding to year-on-year growth of 86%, 102%, and 83%. The bank said the new model includes demand for CPU and optical module PCBs, and for the first time provides a forecast for 2028.

Citibank expects that in 2027, AI-PCB demand will be led by ASICs, accounting for about 34%. Demand related to NVIDIA GPUs will account for 24%, CPUs for 16%, switches for 14%, and optical modules for 12%. In terms of growth rates, network-related demand is expected to be the strongest—especially as 1.6T optical modules and switches ramp up in volume—which could drive optical module PCB growth of 135% in 2026 and 178% in 2027.

The report states that PCB procurement related to Google TPU could become a key incremental driver in 2028. Citibank estimates that Google TPU PCB procurement could reach $16 billion by 2028, exceeding NVIDIA GPU-related PCB demand. Based on the breakdown of incremental demand added from 2027 to 2028, Google TPU contributes about 30%, NVIDIA about 25%, and switches and optical modules each contribute 14%.

As AI-PCB demand expands rapidly, capacity may become a new constraint. Citibank expects that the next round of PCB capacity expansion announcements may begin in the second half of 2026 to support demand in 2028. The bank said that it typically takes 13 to 15 months for PCB factories on the Chinese mainland to go from greenfield projects to mass production, while upstream fiberglass and copper-clad laminate expansion cycles are longer, possibly reaching 18 months. This implies that AI-CCL supply may become even tighter.

Citibank believes that as major AI-PCB projects start in the second half of 2026, first-tier PCB manufacturers with raw-material lock-in capabilities will see improved profit visibility in 2027. Related capacity expansion announcements could also become industry catalysts, strengthening investors’ confidence in the long-term growth of AI-PCBs.

For individual stocks, Citibank maintains “Buy” ratings for Huatian Technology and Shenghong Technology. It raised Huatian Technology’s target price from 119 yuan to 189 yuan, citing the company’s strong mass-production and delivery capabilities in data center switch and AI server PCB areas. Citibank expects Huatian Technology’s net profit to be 6.7 billion yuan, 12.1 billion yuan, and 23.2 billion yuan from 2026 to 2028, respectively, and believes its earnings CAGR could reach 86%.

Shenghong Technology’s target price was raised from 415 yuan to 456 yuan. Citibank lowered its 2026 net profit forecast, mainly reflecting that Rubin-related revenue contribution is below earlier expectations. However, it believes profits will be broadly stable in 2027 and, for the first time, introduced a 2028 net profit forecast of 23.2 billion yuan. The bank said that Shenghong Technology’s long-term relationship with NVIDIA, its HDI capacity, and potential opportunities in data center switches and ASICs still support its valuation premium.

That said, Citibank also cautioned that the AI-PCB market still faces multiple risks, including GenAI-related order share lower than expected, yield problems, price competition in the automotive supply chain, cloud providers cutting capital expenditures, weakening macro demand, rising material costs, and China-U.S. geopolitical risks.

Overall, Citibank’s view is that AI-PCB demand is no longer solely driven by NVIDIA GPU as a single main line, and is expanding into ASICs, Google TPUs, high-speed switches, and optical modules. As demand for 2028 moves ahead into the capacity-planning period, leading Chinese PCB manufacturers with advanced process capabilities, customer resources, and material assurance are likely to continue to be core beneficiaries of AI infrastructure expansion.