Shares of AI-focused printed circuit board (PCB) manufacturers surged more than 9% in Hong Kong trading after NVIDIA Corp. CFO Colette Kress confirmed that the company’s next-generation Rubin architecture is on track for shipments in the second half of 2026, signaling a major demand driver for the sector.
"The demand for GB300 and BL72 was particularly strong, with frontier model builders and hyperscalers each having cumulatively deployed hundreds and thousands of Blackwell GPUs, marking the fastest product ramp in our company’s history," NVIDIA CFO Colette Kress said on the company's recent earnings call. "Brace Blackwell is the fastest training system as well as the lowest token generation cost at inference.”
KB Laminates (1888.HK) jumped 9.1% to a record high of HKD49.32, while the recently listed VGT (2476.HK) soared 9.1% to HKD377.8. The broader Hang Seng Index added 210 points, or 0.8%, to close at 25,596. The rally was fueled by NVIDIA's disclosure that its Rubin and Rubin Ultra platforms will replace traditional copper cables with PCB mid-boards and orthogonal backplanes, a technical shift that requires more sophisticated and higher-value components.
The move is expected to trigger a significant upgrade cycle across the PCB industry, benefiting suppliers capable of producing the high-end boards required for next-generation AI systems. Sinolink Securities noted in a recent report that many AI-PCB companies are already operating at full capacity with strong order backlogs, and are actively expanding production to meet the wave of new demand from AI hardware providers.
NVIDIA's Roadmap Drives Supply Chain Rally
The confirmation of the Rubin timeline during NVIDIA's Q1 FY2027 earnings call provided a clear catalyst for the PCB sector. The company posted record quarterly revenue of $81.6 billion, a staggering 85% year-over-year increase, driven by its data center business.
NVIDIA's CEO Jensen Huang has been clear that the company's one-year-old cadence for new chip architectures is a response to the blistering pace of AI development. The upcoming Vera Rubin platform, which pairs the Rubin GPU with the new Vera CPU, is already seeing strong demand from major AI firms. NVIDIA expects its new Vera CPU platform to generate nearly $20 billion in standalone CPU revenue this year, a new market for the company.
Why It Matters: The PCB Bottleneck
For years, advanced packaging has been a critical bottleneck in the AI supply chain. NVIDIA's aggressive production schedule has pushed its primary foundry partner, TSMC, to outsource some of its advanced packaging work for the first time. TSMC is now partnering with Amkor Technology at a new $7 billion facility in Arizona to handle the overflow, a move that underscores the immense demand for AI hardware.
The shift to more complex PCB designs in the Rubin architecture places even greater emphasis on the capabilities of suppliers like KB Laminates and VGT. As AI models become more powerful, the underlying hardware infrastructure, including the PCBs that connect all the components, must evolve to handle higher speeds and data volumes. This technological shift is creating a durable tailwind for the entire PCB supply chain.
This article is for informational purposes only and does not constitute investment advice.
