Tianfeng Securities Guo Mingqi: Nvidia's next-generation Rubin platform begins testing of new materials, with the PCB upgrade cycle approaching.

NVIDIA’s next-generation AI server platform supply chain deployment is accelerating, and the outline of a new PCB material upgrade cycle is gradually becoming clear.

According to recent supply chain research by Tianfeng International Securities analyst Guo Mingchi, NVIDIA has started testing the next-generation copper-clad laminate (CCL) material M10 with PCB manufacturers, targeting applications including orthogonal backplanes and exchange blade motherboards for Rubin Ultra and Feynman platforms.

If the testing progresses smoothly, M10 CCL and PCBs are expected to enter mass production in the second half of 2027.

This M10 testing also involves two CCL suppliers, breaking the previous reliance on a single supplier, Elite Material (Taiwan Gaoxin), for M9 materials, potentially increasing NVIDIA’s supply chain management flexibility for CCL.

M10 Testing Initiation

According to Guo Mingchi’s supply chain investigation, NVIDIA and Huatian Electronics have officially begun testing M10 CCL materials. Sampling started in the first quarter of 2026, with preliminary results expected in the second quarter of 2026.

The main application scenarios for M10 include two categories: one is orthogonal backplanes (mid-plane) designed to replace existing socket architectures, and the other is exchange blade motherboards for Rubin Ultra and Feynman platforms.

Both application types are central to NVIDIA’s next-generation AI server architecture, with significantly higher performance requirements for PCB materials compared to current solutions.

Guo Mingchi believes that the progress of these tests indicates that Huatian Electronics has gained a leading position in the development of PCBs for Kyber racks and Rubin Ultra and Feynman platforms, which will support the company’s future growth momentum.

Changes in Supplier Landscape

The M10 testing marks a significant change in the supplier structure. In the previous M9 material testing, only Taiwan Gaoxin was qualified as a CCL supplier.

In the M10 phase, NVIDIA has expanded the testing scope to three suppliers, adding two Chinese manufacturers besides Taiwan Gaoxin.

The diversification of suppliers suggests that NVIDIA’s bargaining power and supply chain resilience for high-end CCL procurement are likely to improve, while also creating a more direct competitive environment among related suppliers.

Regarding the technological evolution of the materials themselves, Guo Mingchi pointed out that from a manufacturability and commercialization perspective, the quartz fabric used in M10 could potentially be replaced by Low Dk-2 glass fiber, which may further impact material cost structures and supplier qualifications.

If the M10 testing proceeds as scheduled, with mass production targeted for the second half of 2027, a new cycle of large-scale procurement for AI server PCB materials will begin, providing potential performance catalysts for related supply chain companies.

Risk Warning and Disclaimer

Market risks exist; investments should be cautious. This article does not constitute personal investment advice and does not consider individual users’ specific investment goals, financial situations, or needs. Users should consider whether any opinions, viewpoints, or conclusions in this article are suitable for their particular circumstances. Invest at your own risk.