Even with a record $56 billion planned for capital expenditures in 2026, the world’s largest chip foundry says it cannot build capacity fast enough to meet the relentless demand for artificial intelligence hardware.
Back
Even with a record $56 billion planned for capital expenditures in 2026, the world’s largest chip foundry says it cannot build capacity fast enough to meet the relentless demand for artificial intelligence hardware.

Taiwan Semiconductor Manufacturing Co. is pouring a record $56 billion into capital spending for 2026, yet the company forecasts that the global shortage of advanced AI chips will extend beyond 2027. The capacity crunch at the world’s dominant chipmaker signals a persistent bottleneck for tech giants like Nvidia, Apple, and AMD, who are all reliant on TSMC’s cutting-edge manufacturing to power their next-generation products.
"Our conviction in the multiyear AI megatrend remains high, and we believe the demand for semiconductors will continue to be very fundamental," C.C. Wei, chief executive of TSMC, said during the company's recent earnings call. "Based on our assessment, to meet the strong demand for AI applications, we are stepping up our CapEx investment to increase our N3 capacity."
The spending plan follows a blockbuster first quarter where revenue surged 40.6% year-over-year to $35.9 billion, driven by demand for its most advanced process technologies. High-performance computing (HPC), a category that includes AI accelerators, accounted for 61% of Q1 revenue. Chips made on advanced 3-nanometer (3nm) and 5-nanometer nodes, which are essential for high-end GPUs, made up 25% and 36% of revenue, respectively.
The sustained shortage underscores the massive lead times and capital intensity of semiconductor manufacturing, creating a structural headwind for the entire AI sector. While the investment is a boon for equipment suppliers, it means that for the next two to three years, the growth of AI services could be constrained not by software or algorithms, but by the physical availability of silicon. This dynamic could impact earnings for hyperscalers and chip designers who are unable to secure sufficient supply.
To address the supply-demand imbalance, TSMC is executing an ambitious global expansion focused on its 3nm process, the technology used to produce the most powerful AI chips. The plan includes a new fab in Tainan, Taiwan, scheduled for volume production in the first half of 2027, and a second fab in Arizona set to come online in the second half of 2027. A third facility in Japan is now also slated for an upgrade to 3nm, with production expected in 2028.
Despite these efforts, CEO C.C. Wei acknowledged the challenge. "We try very hard to speed it up and pull in all the equipment as we can, [but] our supply is very tight," he said. The bottleneck extends beyond just wafer fabrication to advanced packaging technologies like CoWoS (Chip-on-Wafer-on-Substrate), which is critical for assembling complex AI accelerators. Goldman Sachs analysts project TSMC’s CoWoS capacity will need to grow 95% in 2027 just to keep pace.
The prolonged shortage is forcing some of the world’s largest tech companies to diversify their supply chains. Tesla is reportedly working with both TSMC and Samsung for its next-generation AI chips, while also partnering with Intel on its "Terafab" initiative. Intel, for its part, is hoping to win major customers with its upcoming 14A process.
The pressure is also being felt in the memory sector, a critical component of AI systems. Major producers like SK Hynix and Samsung are prioritizing high-bandwidth memory (HBM) but are still expected to meet only 60% of AI-related demand by 2027, according to a report from Nikkei Asia. This parallel shortage in a key adjacent market complicates the supply chain further for companies like Nvidia, which became TSMC’s largest customer in 2025, accounting for 19% of its revenue.
For investors, TSMC's situation highlights both opportunity and risk. The company's stock has surged 142% over the past year, reflecting its indispensable role in the AI boom. Analyst price targets from firms like Barclays and Needham have been repeatedly raised. However, the inability to meet demand, even with unprecedented spending, reveals a fragile supply chain that could cap the growth of the entire technology sector.
This article is for informational purposes only and does not constitute investment advice.