The artificial intelligence (AI) boom is reshaping Taiwan’s printed circuit board (PCB) and IC substrate industries, creating unprecedented growth opportunities while exposing critical supply chain bottlenecks. Industry leaders examined the opportunities and risks ahead at the Leaders’ Summit of TPCA Show 2026, held at the Taipei Nangang Exhibition Center on October 21.

Driven by surging demand for advanced materials and increasingly complex architectures in AI data centers, the Taiwan Printed Circuit Association (TPCA) forecasts that the combined production value of Taiwanese PCB manufacturers worldwide will reach about US$35.7 billion in 2026, surpassing the NT$1 trillion (about US$31.3 billion) milestone for the first time.

Despite an optimistic growth outlook through 2030, manufacturers face mounting challenges in scaling production efficiently. These include increasingly complex manufacturing processes, power management and thermal constraints, shortages of critical materials, prolonged equipment lead times, and growing difficulties in precision measurement and inspection.

AI accelerates technology cycles, straining production capacity

“The technology upgrade cycle has shortened from five years to just five months, while substrate sizes continue to increase and PCB layer counts have more than doubled,” said Timothy Lan, executive president of Unimicron Technology Corp.

“Only defect-free products can be shipped to customers. Achieving acceptable yields therefore requires greater manufacturing discipline, higher levels of automation, and tighter environmental controls,” Lan added.

Unimicron Chairman S.C. Chien, who also chairs TPCA, underscored the unprecedented pace of AI development during the show’s opening ceremony. He noted that the doubling cycle for AI computing power has accelerated from approximately two years to just five months since 2024, with demand for XPUs projected to reach six times its 2024 level by 2028.

Despite aggressive capacity expansion by Taiwanese manufacturers, AI substrate production faces a dual bottleneck in materials and equipment. According to Chien, supplies of key materials are falling roughly 20% short of demand, while lead times for certain specialized manufacturing equipment have stretched beyond 30 months.

The shortages are fueling growing frustration among PCB manufacturers and their customers, particularly over constrained supplies of critical materials such as T-glass fabric and advanced copper foil, as well as lengthy equipment delivery schedules.

Frank Liang, vice chairman and CEO of Gallent Precision Machining Co., a PCB and semiconductor equipment supplier, argued that the multifaceted challenges can no longer be addressed simply by increasing capacity.

Instead, he called for closer collaboration across the entire semiconductor and PCB ecosystem to develop integrated advanced packaging solutions. Coordinated optimization and innovation, he said, are essential to reducing the industry’s dependence on a handful of established suppliers and overcoming structural bottlenecks.

His remarks highlight a broader challenge: as AI hardware architectures become more sophisticated, manufacturing constraints increasingly extend beyond individual components or suppliers, requiring coordinated solutions across materials, equipment, processes, and packaging technologies.

AI drives unprecedented growth in high-end PCBs and substrates

Extraordinary circumstances create extraordinary opportunities.

After two decades of modest annual growth of less than 3% between 2000 and 2020, Taiwan’s PCB and substrate industries have entered a new expansion cycle fueled by the generative AI boom that began in 2024.

Shiuh-kao Chiang, managing partner of Prismark Partners LLC and a veteran PCB industry analyst, projects that global production value for high-end PCBs will reach about US$3.4 billion in 2026, representing a 32% year-on-year increase. He also expects the IC substrate segment to record a 40% surge in production value this year, setting another industry record.

The sharp acceleration reflects a fundamental shift in demand. AI infrastructure requires larger, more sophisticated, and higher-value PCBs and substrates, creating opportunities for manufacturers capable of meeting increasingly stringent technical and quality requirements.

However, the industry’s rapid expansion also raises questions about how much of the current demand is sustainable.

Three risks threaten the sustainability of the AI boom

Elaine Huang, deputy editor-in-chief of CommonWealth Magazine, who moderated the panel discussion, cautioned that three major risks deserve attention amid the industry’s optimism.

The first is the potential crowding out of consumer electronics as AI infrastructure commands an increasing share of manufacturing capacity and critical components, leaving consumer electronics and other non-AI applications struggling to secure adequate supplies.

The second is demand front-loading or double-booking, as customers rush to secure supplies amid fears of future shortages. Such precautionary ordering could inflate near-term demand and create the risk of an inventory correction later.

PC manufacturers, for example, are concerned that aggressive advance ordering could effectively pull forward demand from the second half of the year, leaving a weaker market once inventories have been replenished.

The third risk concerns the sustainability of massive capital expenditure and the availability of sufficient power infrastructure to support continued AI expansion.

Together, these risks suggest that the PCB industry’s growth trajectory will depend not only on technological advances and capacity expansion but also on the underlying strength and sustainability of AI infrastructure investment.

For Taiwan’s PCB and substrate manufacturers, the challenge is therefore twofold: scaling production fast enough to capture unprecedented AI demand while avoiding overexpansion in response to potentially inflated short-term orders.

The AI boom has opened a historic growth window for Taiwan’s PCB industry. But as technology cycles accelerate and supply chain constraints deepen, the companies best positioned to benefit may not be those that expand capacity the fastest, but those that can combine manufacturing innovation, supply chain resilience, and disciplined investment to sustain growth beyond the current surge.

Have you read?

• Taiwan’s Auto Parts Makers Need Reliability Data to Win in AI Cooling
• Taiwan Builds World-Class Products. It Forgets to Tell Anyone
• How Taiwan Can Build Global Brands Without Losing Itself

Uploaded by Ian Huang