[News] TSMC Equipment Demand Reportedly Jumps 90% in Just Over Half a Year; Substrate Capacity Faces AI Squeeze

AI demand is pushing TSMC’s capacity requirements to unprecedented levels. According to Liberty Times, TSMC Senior Vice President and Co-COO Cliff Hou said he has never seen demand grow at such a rapid pace and frequency in the past 30 years. Hou illustrated the surge through TSMC’s equipment demand: equipment demand stood at 1x at the end of last year and climbed to 1.9x by July. That represents a 90% increase in just over half a year, a pace Hou said poses a major challenge for Taiwan’s semiconductor ecosystem.

Technology is another challenge. As noted by Economic Daily News, Hou said AI’s growing appetite for computing power and energy efficiency is accelerating demand for advanced technologies, advanced packaging, and system-level performance far faster than before, requiring a new model of industry collaboration.

Improving individual chip performance alone is no longer enough, Hou said. Semiconductor production was once more linear, with each participant primarily ensuring its product could move successfully to the next manufacturing stage. Today, companies must ensure that performance achieved at one stage can be fully utilized by the next partner while maintaining manufacturability and yield, requiring much closer supply-chain collaboration, Economic Daily News highlights.

To address the capacity bottleneck, TSMC is building fabs on an unprecedented scale. According to Liberty Times, Hou said the company is simultaneously constructing 13 fabs in Taiwan and another five to six worldwide, bringing the total to nearly 20. TSMC previously had only four or five fabs under construction at a time, yet Hou said even the current expansion is “still not enough to meet demand.”

AI Demand Strains Substrate Capacity and Supply Chain

Meanwhile, at the same event, Unimicron Chairman S.C. Chien highlighted severe capacity constraints in the semiconductor substrate market. According to TechNews, surging AI demand has intensified the supply-demand imbalance, while requirements for larger substrates, higher layer counts, and more complex designs—including lower coefficients of thermal expansion (CTE) and dielectric constants—are raising technical barriers and manufacturing complexity.

Chien said the substrate supply chain is at a critical juncture, with key production tools and materials, including ABF and TGV-related technologies, heavily reliant on small and medium-sized Japanese suppliers. Over the past 18 months, Unimicron has held more than a dozen in-depth meetings with customers and Japanese suppliers.

Beyond the race for advanced-process capacity, mature nodes also remain essential in the AI era. TechNews highlights Chien’s view that advanced and mature technologies are highly complementary within the AI ecosystem. He said emerging AI applications are creating new opportunities for mature nodes, particularly in peripheral and sensing applications requiring customized solutions. Chien pointed to high-density bonding, specialty memory, logic-based dies, silicon interposers with deep-trench capacitors, and silicon photonics as areas with significant potential for mature node.

Read more

• [News] TSMC Eyes Chip-Level Microchannel Cooling as AI System Power Could Rise 6× in Five Years
• [News] TSMC: SoIC, CoWoS to Drive 50× Compute by 2029; Silicon Photonics to Top 50% of Transceiver Market by 2027

(Photo credit: TrendForce)