Journal · 2026-10-30 · 10 min read

PCB Shortages Extend to 2030

PCB supply shortages, driven by AI demand, are now projected to last until 2030, with lead times stretching to 30 months.

Printed circuit boards (PCBs) are facing a severe, long-term shortage that will push up prices for motherboards, GPUs, and other PC components well into the 2030s. According to Nikkei Asia, orders placed today may not be fulfilled until late 2028, with some PCB designs taking up to 30 months to deliver. This crisis is fueled by the AI industry consuming nearly all available manufacturing capacity, leading to skyrocketing material costs and expanded lead times. Copper-clad laminates (CCLs), a critical PCB raw material, have seen prices surge 12 times since July 2026. NVIDIA’s partnerships with chip substrate makers like Ibiden, Samsung Electro-Mechanics, and Unimicron are accelerating demand for advanced substrates, further straining the supply chain. As a result, PC makers are struggling to secure enough PCBs despite declining PC shipments, and the entire electronics ecosystem is bracing for another round of price hikes.

PCBs at the Heart of the Crisis

Printed circuit boards (PCBs) are the foundational layer of every electronic device, serving as the physical and electrical backbone that connects chips, memory, power circuits, and peripherals. They are constructed using layers of copper, fiberglass, and insulating materials, with signal traces etched into the copper to route electrical signals. In modern systems, high-end motherboards and GPUs use up to 20 layers, with complex routing for high-speed data and power delivery. The increasing complexity of AI systems—especially data center servers—has made PCBs a critical bottleneck. These systems require PCBs with high copper content, specialized fiberglass for thermal stability, and dense interconnects for memory and processor clusters. As a result, the demand for PCBs has surged, pushing manufacturers to their limits.

The AI-Driven Supply Chain Crunch

The AI industry is now the primary driver of PCB shortages, consuming nearly all available production capacity. According to Tom’s Hardware, AI data centers are “gobbling up” manufacturing resources that were once dedicated to PCs, smart devices, and consumer gadgets. This shift has caused a ripple effect: traditional PC and consumer electronics demand is being underserved as AI products—especially high-performance GPUs and accelerators—command higher margins and priority access to materials and fabrication. As a result, PCB makers are prioritizing orders for AI servers and AI chips, leaving less capacity for standard PC components. The combination of rising demand and limited supply has created a tight market where lead times have doubled and delivery delays are now the norm.

Material Costs Surge: CCLs and Fiberglass

A key factor in the PCB shortage is the dramatic increase in the cost of raw materials. Copper-clad laminates (CCLs), the core material used in PCB construction, have seen prices rise 12 times since July 2026, according to a Taiwanese PC maker. This surge is due to increased demand from both AI systems and the broader semiconductor industry. The specialized fiberglass used in PCBs—critical for thermal stability and signal integrity—has also become harder to source. One of the world’s leading CCL suppliers reported to Nikkei that its order book is already filled through 2030, despite plans to triple capacity by 2028. This indicates that the supply chain is not only strained but also in a state of long-term commitment, with new orders locked in for years.

Lead Times Stretch to 30 Months

The most striking metric in the current PCB crisis is the lead time for orders. According to Tom’s Hardware, some PCB designs now require 30 months from order placement to delivery. An unnamed American PC maker executive confirmed this, stating that even with declining PC shipments, the industry is “still struggling to secure enough PCB supplies to meet our needs.” These lead times are driven by multiple factors: capacity constraints in PCB fabrication, limited availability of CCLs and fiberglass, and a shortage of skilled engineering staff. The 30-month timeline means that a motherboard or GPU design ordered in October 2026 will not be delivered until 2029, making long-term planning essential for PC manufacturers.

NVIDIA’s Strategic Moves in the Substrate Space

NVIDIA has been at the forefront of addressing the PCB and substrate shortage through strategic partnerships. The company recently announced collaborations with three major chip substrate makers: Japan’s Ibiden, South Korea’s Samsung Electro-Mechanics, and Taiwan’s Unimicron. Chip substrates are specialized PCBs that serve as the base for mounting silicon dies in CPUs and GPUs. These substrates are critical for high-performance chips, especially those using CoWoS (Chip-on-Wafer-on-Substrate) or HBM (High Bandwidth Memory) stacking. Unimicron is now planning a 2.5x expansion of its substrate capacity over the next three years, aligning with NVIDIA’s roadmap for next-generation AI accelerators. These partnerships ensure that NVIDIA’s high-end GPUs, such as the RTX 5090, will have priority access to advanced substrates, which in turn strengthens the demand for PCBs across the ecosystem.

The Human Factor: Engineering Talent Shortage

Beyond materials and equipment, the PCB industry is facing a critical shortage of skilled engineering staff. According to Tom’s Hardware, salaries across the PCB sector have risen 20% to 30% in recent months to retain talent. This increase reflects the growing complexity of PCB design and manufacturing, particularly for AI-optimized boards. Engineers now need expertise in high-speed signal integrity, thermal management, and advanced layering techniques. The demand for these specialists is so high that companies are racing to establish new production facilities, but they struggle to staff them. As one PCB supply firm noted, it is increasingly difficult to hire and retain engineers for new production lines, making workforce development a key challenge in the supply chain.

Price Hikes Across the PC Ecosystem

The PCB shortage is not a single event but a cascading effect across multiple product lines. TechPowerUp reports that AI server demand is consuming all PCB manufacturing capacity, leading to another round of price hikes across the PC ecosystem. This includes motherboards, GPUs, and even SSDs and RAM, which are also affected by the broader semiconductor crisis. The rise in CCL costs alone has contributed to a 12x increase in material expenses, which are now being passed on to consumers. As a result, PC makers are forced to raise prices for their products, with the median price of NVIDIA’s RTX 50 series jumping 41% in August 2026. The MSRP for these GPUs is now considered a “distant dream,” as supply chain bottlenecks push actual retail prices significantly higher.

Who Is Affected?

The impact of the PCB shortage is widespread across multiple sectors. Hyperscalers like Google, Microsoft, and Amazon are prioritizing PCBs for their AI data centers, ensuring high-end servers but leaving smaller players with limited access. Cloud renters—especially AI startups and research labs—are now facing higher costs for GPU instances and longer provisioning times. Small studios and AI developers are adjusting their budgets to account for delayed hardware deliveries and increased component prices. Hobbyists, who once relied on affordable, off-the-shelf components, are now paying premiums for motherboards and GPUs, with some even building systems around older DDR3 technology to reduce dependency on scarce components. The crisis has also prompted a wave of community-driven workarounds, such as modding existing GPUs to improve performance and reduce power consumption.

Alternatives and Workarounds

To mitigate the PCB shortage, several workarounds have emerged. One approach is the use of hybrid PCB designs that combine standard and high-performance layers, optimizing cost and performance. Another is the adoption of modular systems where components can be upgraded independently, reducing the need for full motherboard replacements. Modders are also enhancing older GPUs—such as RTX 20, 30, and 40 series—with custom firmware and cooling solutions, boosting compute performance by up to 2x. These efforts, highlighted in a GeForce Unlocked article from October 2026, allow users to extend the life of existing hardware while waiting for new PCB-based components. Additionally, some manufacturers are offering bundled packages that include PCBs, substrates, and memory modules to reduce dependency on external suppliers.

Counter-Arguments and Risks

Despite the clear signs of a deepening crisis, some analysts question whether the current trajectory is sustainable. According to TechPowerUp, the 30-month lead time is a forecast based on current orders, but if demand continues to grow, lead times could extend beyond 2030. Tom’s Hardware notes that while PCB makers are investing in expansion, the pace of new facility ramp-up is slower than expected. The risk is that the industry will fall behind, leading to a backlog of unfulfilled orders even after 2030. Furthermore, the 12x increase in CCL costs may not be fully recoverable if copper prices stabilize, leaving manufacturers with margin pressure. Another risk is the potential for supply chain disruptions due to geopolitical factors or natural disasters, which could delay material deliveries and further stretch lead times.

What We Still Don’t Know

Several key questions remain unanswered. First, the exact composition of the PCB order books is unclear. While Nikkei reports that one CCL supplier’s book is filled through 2030, the breakdown of orders by product type (e.g., AI servers vs. consumer GPUs) is not detailed. Second, the impact of the 20% price hike across AMD’s GPUs, chipsets, and CPUs—announced in September 2026—on the overall PCB supply chain is not yet quantified. Third, the full scope of NVIDIA’s partnerships with Ibiden, Samsung Electro-Mechanics, and Unimicron remains to be explored, including how these collaborations affect the availability of substrates for non-NVIDIA chips. Finally, the long-term effects of the 20–30% salary increases on engineering talent are still uncertain, particularly regarding retention and productivity.

What to Watch and When

The next critical milestones to monitor include: the delivery of the first batch of PCBs ordered in October 2026, expected by late 2028; the completion of Unimicron’s 2.5x capacity expansion by 2029; and the finalization of the CCL supplier’s 2030 order book by Q1 2027. The 2027 AI hardware launch season will be a key indicator of whether the supply chain can keep pace with demand. Additionally, the performance of modded RTX 20–40 series GPUs in real-world AI workloads, as reported in the GeForce Unlocked article, will shape future expectations. Finally, the results of the upcoming NVIDIA and AMD product launches—especially the RTX 5090 and the next-gen Zen 6 CPUs—will reveal how well the PCB supply chain is supporting high-end hardware. The entire ecosystem will be watching the 2028 deadline closely, as it marks the first major milestone in the long-term PCB shortage.

Summary of Reported Figures

Key Quotes from the Sources

• "Even though PC shipments could decline again next year, we are still struggling to secure enough PCB supplies to meet our needs." — American PC maker executive, Nikkei Asia
• "The AI industry continues to gobble up manufacturing resources that were once dedicated to serving PC, smart device, and gadget makers." — Tom’s Hardware
• "Costs for CCLs have surged at least 12 times since July." — Taiwanese PC maker
• "A new order today would not be fulfilled until late in 2028." — PCB supply firm, Nikkei Asia
• "Salaries have already risen 20% to 30% across the PCB industry to keep skilled employees." — Tom’s Hardware

Our Reading: The Buyer’s Perspective

For small studios and hobbyists, the PCB shortage is not just a logistical headache but a financial burden. With motherboards and GPUs now taking over two years to arrive, budgeting becomes a strategic exercise. The 41% median price increase for RTX 50 series GPUs means that a $1,000 card may cost $1,410 in practice, forcing buyers to reconsider their hardware choices. The modding community’s response—boosting older GPUs by up to 2x—offers a viable alternative, but it requires time, skill, and investment. As a result, buyers are becoming more selective, prioritizing longevity and modularity over raw performance. The current landscape favors those who can afford to plan ahead, while others are left scrambling for components that arrive months—or even years—late.

← All notes