Journal · 2026-10-30 · 12 min read

PCB Shortages to Last Until 2030

PCB shortages, driven by AI demand, are set to extend into 2030, with lead times stretching to 30 months and material costs surging 12-fold.

Printed circuit boards (PCBs) are facing a crisis that will ripple across the entire PC ecosystem, with shortages projected to last until 2030. According to Nikkei Asia, an unnamed American PC maker executive confirmed that PCB orders placed today may not be fulfilled until late 2028, and some PCB designs now require 30 months to complete. The AI industry is the primary driver, consuming vast amounts of PCB manufacturing capacity, copper-clad laminates (CCLs), and specialized fiberglass. As a result, PCB prices are set to rise further, affecting motherboards, GPUs, and other components. CCL prices have increased 12 times since July 2026, and leading suppliers have order books filled through 2030, despite plans to triple capacity by 2028. NVIDIA’s recent partnerships with chip substrate makers like Ibiden, Samsung Electro-Mechanics, and Unimicron further highlight the strategic shift toward advanced PCBs. The supply chain is now under strain not only in materials but also in engineering labor, with salaries rising 20% to 30% across the sector. These developments signal a transformation in how PC makers and AI companies view the humble PCB.

PCBs at the Heart of the AI Supply Chain

Printed circuit boards (PCBs) are the backbone of every electronic device, serving as the physical platform where components like processors, memory, and power circuits are mounted and interconnected. In the modern PC, a PCB integrates the CPU, GPU, RAM, storage, and I/O interfaces through intricate copper traces and soldered components. As AI systems grow more complex, they demand multi-layered PCBs with high-density interconnects, fine-pitch traces, and advanced thermal management. These requirements push the limits of traditional PCB manufacturing, requiring more copper, specialized fiberglass, and precise alignment of components. The shift from consumer electronics to AI-driven data centers has elevated PCBs from a supporting role to a central bottleneck in the supply chain.

The Technical Challenge: CoWoS, Interposers, and HBM Stacking

Background: CoWoS (Chip on Wafer on Substrate) is a packaging technology used to stack multiple silicon dies—such as GPUs or AI accelerators—onto a single substrate. The interposer, a thin silicon or organic layer, sits between the dies and the PCB, enabling high-speed communication and power delivery. HBM (High Bandwidth Memory) stacking involves vertically stacking memory dies and connecting them via through-silicon vias (TSVs), drastically increasing memory bandwidth. Together, these technologies require PCBs with extremely high layer counts, fine-pitch routing, and robust thermal pathways. The result is a tightly integrated system where the PCB is no longer just a passive connector but an active participant in performance.

Advanced PCBs for AI workloads must handle power delivery across multiple layers, manage heat from high-density chips, and support signal integrity at GHz frequencies. The use of HBM stacks in GPUs and AI chips demands PCBs with precise alignment and high thermal conductivity to prevent hot spots. As a result, PCBs for AI servers and high-end GPUs now include 20 to 30 layers, with embedded passive components and micro-vias. These advanced PCBs are significantly more complex and costly than those used in consumer laptops or desktops.

The Money Flow: Who Pays and How Prices Move

According to Tom's Hardware, AI data centers are now the dominant consumers of PCBs, driving up prices and reshaping the supply chain. The shift means that PC makers must compete with hyperscalers like Google, Microsoft, and Amazon for limited PCB production capacity. As a result, PC manufacturers are forced to pay higher premiums for PCBs, especially for advanced designs. The cost of copper-clad laminates (CCLs), a key raw material, has surged 12 times since July 2026, directly impacting PCB manufacturing costs. This increase is passed down through the chain, affecting component suppliers, OEMs, and final product prices.

NVIDIA’s partnerships with chip substrate makers—such as Ibiden (Japan), Samsung Electro-Mechanics (South Korea), and Unimicron (Taiwan)—are strategic moves to secure supply. These chip substrates serve as the base for mounting silicon dies in GPUs and CPUs, effectively acting as miniature PCBs. By locking in long-term contracts, NVIDIA ensures steady supply of high-quality substrates, while also raising the bar for PCB quality across the industry. The increased demand for substrates has led to tighter supply, pushing up prices and lead times for both substrates and final PCBs.

Capacity, Lead Times, and Equipment Constraints

Several sources report that PCB order fulfillment times are now stretching to 30 months, with some designs taking even longer. According to TechPowerUp, PCB manufacturers are confirming that orders placed in October 2026 will not be completed until late 2028. This 30-month lead time is due to a combination of equipment bottlenecks, material scarcity, and labor shortages. The expansion of PCB fabrication lines requires significant capital investment, and new production lines are taking over a year to commission.

A key constraint is the availability of specialized fiberglass used in PCB construction. This material must meet strict thermal and electrical specifications, and its production is limited to a few global suppliers. As AI demand grows, these suppliers are prioritizing orders for high-end PCBs, leaving consumer-grade PCBs behind. Additionally, the equipment needed to process high-layer-count PCBs—such as laser drilling machines and high-precision lamination presses—is in limited supply, further extending lead times.

Quotes from the Industry: What Companies Are Saying

According to TechPowerUp, an unnamed executive from an American PC maker told Nikkei Asia: "Even though PC shipments could decline again next year, we are still struggling to secure enough PCB supplies to meet our needs." This statement highlights the tension between declining PC demand and rising PCB scarcity. The executive emphasized that securing PCBs is now a top strategic priority, despite the risk of overstocking in a shrinking market.

Tom's Hardware adds that "a significant number of products have lead times that have doubled," citing PCB makers who report increased complexity and longer production cycles. One PCB supply firm explained to Nikkei that a new order today would not be fulfilled until late in 2028. The same source noted that "the AI industry continues to gobble up manufacturing resources that were once dedicated to serving PC, smart device, and gadget makers," illustrating the shift in industry focus.

An unnamed Taiwanese PC maker reported that CCL costs have surged at least 12 times since July 2026. The world’s leading CCL supplier confirmed to Nikkei that their order book is filled through 2030, despite plans to triple capacity by 2028. These quotes reveal a supply chain under pressure, where demand consistently outpaces supply.

Who Is Affected: From Hyperscalers to Hobbyists

Hyperscalers like Google, Microsoft, and Amazon are the primary beneficiaries of the PCB boom, as they require massive numbers of high-performance PCBs for their AI data centers. These companies are able to absorb higher PCB costs due to their scale and long-term contracts. However, the ripple effect reaches smaller players. Cloud renters—such as AI studios, indie game developers, and research labs—now face higher costs for cloud instances, as providers pass on PCB-related price increases to their customers.

Small AI studios, in particular, are affected by longer lead times and higher component costs. A studio planning to launch a new AI model in 2027 must now account for a 30-month PCB lead time, meaning they must place orders by early 2025 to have components ready in time. This long lead time forces them to plan further ahead and manage inventory with greater precision.

Hobbyists and small-scale builders are also impacted. The increased cost of motherboards and GPUs—driven by PCB shortages—means that building a high-end PC now requires a larger upfront investment. As a result, hobbyists may delay upgrades or choose lower-tier components to manage costs. The RTX 5090, for instance, is now a premium product, with power connectors melting at 175°C, as reported by Hardware Unboxed in September 2026.

Alternatives and Workarounds

To cope with the PCB shortage, companies are exploring several alternatives. One approach is to use more standardized PCB designs, reducing the need for custom layouts and lowering production complexity. Another strategy is to increase in-house PCB fabrication, allowing companies to better control lead times and quality. NVIDIA, for example, has begun producing its own chip substrates in-house, reducing dependency on external suppliers.

Workarounds include using older PCB designs with lower layer counts and less copper, which are cheaper and faster to produce. This allows PC makers to offer entry-level products with acceptable performance, even when high-end PCBs are scarce. Some manufacturers are also adopting modular designs, where components like GPUs and CPUs are mounted on separate PCBs that can be swapped in and out, reducing the need for full system integration.

Another workaround is the use of hybrid PCBs that combine organic and silicon-based materials. These hybrid boards offer better thermal performance and signal integrity, making them ideal for AI workloads. The adoption of such designs is expected to accelerate in the next two years, as companies seek to balance cost and performance.

The Counter-Argument: Could This Not Happen?

Despite the strong evidence, some experts question whether the PCB shortage will truly last until 2030. According to TechPowerUp, the current 30-month lead time may be optimistic, given the complexity of the manufacturing process. The report notes that "some predicting that orders placed today may not be fulfilled until late 2028," suggesting that even a 30-month timeline might be tight. The same source adds that "PCB manufacturers are confirming this situation," indicating that the problem is not just a forecast but a current reality.

Tom's Hardware, however, offers a slightly more cautious view. While the AI industry is clearly driving demand, the report notes that "traditional PCB demand is becoming underserved as AI products demand and pay more." This implies that PC makers are not just facing higher prices but also reduced availability of standard PCBs. The counter-argument is that if the PCB shortage persists, PC makers may shift to lower-tier components, reducing the performance of their products and potentially alienating their customer base.

Furthermore, the success of the current expansion plans depends on timely delivery of equipment and skilled labor. If equipment delivery is delayed or if engineering staff cannot be hired fast enough, the projected capacity increases may fall short, leading to a continued bottleneck. The 12-fold increase in CCL costs since July 2026 is a strong signal, but it remains to be seen whether suppliers can scale production to meet demand.

What We Still Don’t Know

Several key questions remain unanswered. First, the exact breakdown of PCB types and their respective lead times is unclear. While 30 months is the average, some designs may take longer. Second, the impact of the 12-fold CCL price increase on final product costs is still uncertain. How much of this cost is passed to the end user, and how much is absorbed by OEMs?

Another unknown is the role of emerging markets in PCB demand. Are countries like India and Brazil beginning to contribute to the global PCB supply chain, or are they primarily consumers? Additionally, the long-term effects of the 20% to 30% salary increases for engineering staff are not yet clear. Will these higher salaries attract more talent, or will they drive up overall production costs?

Finally, the success of NVIDIA’s partnerships with Ibiden, Samsung Electro-Mechanics, and Unimicron will be crucial. How will these collaborations affect the quality, cost, and availability of chip substrates? Will Unimicron’s plan to expand capacity by 2.5x over three years prove sufficient to meet rising demand?

What to Watch and When

The next six months will be critical. By January 2027, the industry will have a clearer picture of whether the 30-month lead time is sustainable. The first wave of products from the expanded CCL suppliers will arrive, and their quality and cost will be key indicators.

By mid-2027, the impact of the 12-fold CCL price increase on final product pricing will become evident. PC makers will release new product lines, and consumers will begin to feel the effects of PCB shortages. The success of NVIDIA’s chip substrate partnerships will be evaluated, with Unimicron’s expansion progress reported by Q2 2027.

By the end of 2027, the industry will assess whether the order books of CCL suppliers remain filled through 2030. If new orders continue to pour in, the PCB shortage may not only last until 2030 but extend beyond. The 2027 holiday season will be a key test, as PC and GPU sales are expected to rise, putting additional pressure on the supply chain.

Summary of Key Figures from Multiple Outlets

Bullet List: Key Developments in the PCB Industry

• PCB order fulfillment times now stretch to 30 months, with some designs taking longer.
• Copper-clad laminates (CCLs) have increased 12 times in price since July 2026.
• AI data centers are the primary drivers of PCB demand and cost.
• NVIDIA has announced partnerships with Ibiden (Japan), Samsung Electro-Mechanics (South Korea), and Unimicron (Taiwan).
• CCL suppliers have order books filled through 2030, despite plans to triple capacity by 2028.
• PCB makers are investing in new production lines, but equipment and labor are limited.
• Engineering salaries have risen 20% to 30% across the industry.
• Hybrid PCBs combining organic and silicon materials are emerging as a key innovation.
• Workarounds include modular designs, standardized PCBs, and in-house fabrication.
• The 2027 holiday season will be a major test for the supply chain.

Buyer’s Perspective: Our Reading

For small studios and hobbyists, the PCB shortage means a more strategic approach to purchasing. With lead times stretching to 30 months, planning is no longer optional—it’s essential. The increased cost of motherboards and GPUs will force buyers to prioritize performance over features. The melting RTX 5090 power connector, reported in September 2026, is not just a technical issue but a financial one, requiring careful budgeting and system design. As a result, buyers are likely to adopt a more holistic view of their hardware stack, considering not just component performance but also supply chain stability and future scalability.

← All notes