Challenges and Opportunities for PCB Manufacturers in the AI Era

April 17, 2026

AI is red-hot right now.

Whether it’s large AI models or AI assistants, new developments keep emerging one after another, dazzling to watch and constantly stirring the nerves of workers. AI continues to evolve. Whether large models or AI assistants, they all require continuous training, optimization, and iteration—all of which are inseparable from the PCB, often called the “mother of electronic products.” As is well known, PCBs are manufactured by PCB manufacturers through complex processes. Therefore, we can confidently say that PCB manufacturers hold a pivotal position in the PCB manufacturing industry, serving as a bridge between the past and the future.

Bridging the past: PCB manufacturers take various raw materials and transform them through complex manufacturing processes into bare PCBs.
Bridging the future: These bare PCBs then undergo component soldering, assembly, and a series of processes at OEM/ODM/OSAT facilities, ultimately becoming finished products or parts of products in consumers’ hands.

For those familiar with PCB manufacturers, although they play a critical role in the PCB industry, the competition is fierce—especially for PCB manufacturers in China, which are operating in a veritable “red ocean.” Now, into this red ocean comes AI, like a gale-force wind stirring up turbulent waves, prompting a flurry of moves among major PCB manufacturers. Therefore, it is necessary to explore the challenges and opportunities facing PCB manufacturers in the AI era.

AI-Focused PCB Manufacturers: Investing in capacity expansion to consolidate their technological and production advantages in AI-related PCBs. Examples: Shengyi Technology, WUS Printed Circuit, Shennan Circuits.

Leading PCB Manufacturers: Optimistic about the prospects of AI-related PCBs, actively pursuing capacity expansion and product diversification strategies in this area. Examples: Zhen Ding Technology, Unimicron, Kinwong, Sheng Yi Electronics.

Challenges Facing PCB Manufacturers

Pressure from Upstream and Downstream

In the AI era, demand for AI-related PCBs—such as those used in AI servers—has surged. High-layer-count boards (HLC) and high-order HDI (including ultra-HDI) boards have become the new market favorites. However, beneath the surface prosperity lie turbulent undercurrents: structural shortages of high-end materials, low manufacturing yields, a significant gap between R&D investment and talent availability, compounded by downstream customers’ relentless pressure on PCB pricing and ever-stricter quality requirements. All these test the adaptability of PCB manufacturers.

Structural Shortage of High-End Raw Materials: Long lead times, with 180-day lead times being not uncommon.

Direct Raw Material Prices Rising Across the Board: Continuous price increases for substrates (CCL), prepreg (PP), copper foil, and other direct materials directly drive up PCB costs and continuously squeeze PCB manufacturers’ profit margins.

High-End Material Capacity Crowding Out Mid-to-Low-End Capacity: This leads to capacity constraints or production halts, triggering material change processes. The change process (PCN) can take weeks or even months, adding to manufacturers’ costs: testing costs, management costs, capacity pressures, etc.
PCN Reference: PCB Manufacturer Change Management.

Upgrading Indirect Materials Imperative: For example, existing drill bits are inefficient. For M9 materials, drill bit life is only 100–200 holes, making new drill bits essential. Currently, diamond drill bits are already undergoing market testing. PCD (polycrystalline diamond) drill bits, thanks to their extremely high hardness, wear resistance, and processing stability, can achieve a lifespan dozens or even hundreds of times longer than conventional solutions when working with M9-grade materials, while also significantly improving hole wall quality and processing efficiency. The demand logic for diamond drill bits is expected to shift from optional tools to essential processing tools for high-end PCBs, with industrialization verification currently accelerating. The trend toward upgrading PCB material systems is clear, and the penetration rate of diamond drill bits is expected to accelerate rapidly, with the market size entering a growth cycle.

Manufacturing Process Yields

• Lamination Alignment Accuracy: Whether it’s the single lamination of multiple cores for high-layer-count HLC boards or the multiple lamination steps for high-order HDI boards, interlayer alignment accuracy remains a manufacturing challenge.

 





• Drilling Quality: AI-related PCBs often use high-Tg, high-speed, high-frequency, and thick-copper designs, which can easily lead to increased drilling roughness, burrs, and difficulty in desmearing. Additionally, issues such as low drilling efficiency and frequent drill breakage need to be addressed.

• Deep Hole Plating: The quality of deep hole plating for aspect ratios exceeding 20:1 directly determines the electrical performance of the PCB.

 





• Fine Line Quality: High-density designs with fine lines are prone to increased open circuits, short circuits, and micro-shorts, affecting yields.

• Special Processes: Techniques such as back drilling (controlled depth drilling) and VIP (via-in-pad) also add to manufacturing complexity.

 

R&D Investment and Technical Talent

There is considerable room for yield improvement, which requires both financial resources and technical expertise. Based on publicly available information, the current yield for PCBs used in NVIDIA’s CPS Ruby server is around 50%, with expectations to reach 70% in the future. However, improving yields requires not only investment in new equipment and processes but also technical personnel to optimize testing solutions in order to shorten the yield improvement cycle. New plants, new equipment, and new processes all require technical talent to transform process capability into manufacturing capability.

Relentless Pressure from End Customers on PCB Price and Quality

• PCB price reductions are a routine practice for customers—annual, quarterly, or even monthly price cuts.

• Customer PCN processes challenge PCB manufacturers’ change management capabilities and impose additional costs and capacity pressures.

• Rising raw material prices further shrink PCB manufacturers’ pricing flexibility.

Downstream Customers’ Escalating Quality Requirements

• Adopting incoming inspection standards that exceed end-customer requirements.

• Transferring their own KPI and cost pressures to PCB manufacturers.

• Requiring on-site support from PCB manufacturers for abnormality confirmation and handling.

Opportunities for PCB Manufacturers

The advent of the AI era has elevated the PCB industry to an unprecedented strategic height, bringing with it a substantial incremental market for PCB manufacturers—a growth trajectory that is both long-term and certain.

Increase in PCB Value

Taking server PCBs as an example: traditional server PCBs typically use 12–16 layers, whereas AI server mainboards have already increased to 20–28 layers, with some high-end products reaching 34–50 layers or more. This increase in layer count directly drives up product value—the PCB cost per AI server rack has risen sharply, with high-end product prices reaching several times those of traditional servers.

Expansion Boom Focused on High-End Capacity

In response to the market opportunities presented by AI, leading PCB manufacturers are accelerating capacity expansion, with investment heavily concentrated on high-end capacity.

In the AI era, the PCB industry is undergoing a profound restructuring—from individual competition to ecosystem-level collaboration. For PCB manufacturers, while embracing the incremental market, they must also consider how to better seize opportunities and broaden their competitive advantages.

Supply Chain Level: Leveraging their pivotal bridging position, PCB manufacturers need to establish deeper technical collaboration with upstream suppliers of copper foil, fiberglass fabric, and copper-clad laminates, and even secure critical capacity through capital ties.

Customer Relationship Level: Actively collaborate with customers on the development of new products, technologies, and processes, expanding their own technological depth and forging strategic partnerships with customers.

Technology and Talent Level: Actively promote industry-education integration and school-enterprise cooperation to accelerate technology implementation and shorten talent development cycles.

Software Level: Leverage the vast data resources of PCB manufacturers to actively explore PCB simulation and modeling with PCB software and equipment suppliers, improving technology implementation efficiency and reducing manufacturing and testing costs.

Intelligence Level: Build AI-centric smart factories and data backbones to improve PCB manufacturing efficiency and yields.

Conclusion

The PCB industry in the AI era is in a period of incremental market growth, an era where challenges and opportunities coexist. For PCB manufacturers, the pace of capacity implementation directly determines their survival. Therefore, it is my view that the future winners will inevitably be those PCB manufacturers capable of simultaneously navigating multiple transformations:

• Technologically: Achieve continuous breakthroughs in high-layer-count HLC and high-order HDI, constantly advancing manufacturing technology and widening their technological moat.

• Manufacturing-wise: Build AI-centric smart factories, continuously push the boundaries of manufacturing capability, and achieve stable mass production quality.

• Ecosystem-wise: Deeply embed themselves in regional innovation networks and global supply chains to ensure sustained supply chain stability.