Samsung’s much-anticipated semiconductor fabrication plant in Taylor, Texas, is facing fresh delays, with sources attributing the setback to a combination of low customer demand and rapidly shifting chip technology standards.
The $44 billion facility, envisioned as a centerpiece of America’s push to reclaim chip-making supremacy, is now highlighting a critical flaw in Washington’s semiconductor strategy: an overemphasis on supply without addressing weakening demand, worsened by restrictive U.S. export policies.
The South Korean tech giant began constructing the Taylor site in 2022 with a $17 billion investment, later expanding its ambition to a multi-fab campus supported by a $6.6 billion CHIPS Act subsidy. Although the factory is now reportedly 92% complete, sources told Nikkei Asia that Samsung cannot move forward with chip production because it has no guaranteed buyers.
“Local demand for chips isn’t particularly strong, and the process nodes Samsung planned several years ago no longer meet with current customer needs,” the executive told Nikkei Asia. “However, overhauling the plant would be a major and costly undertaking, so the company is adopting a wait-and-see approach for now.”
Originally intended to produce 4nm chips, Samsung is now aiming for 2nm production to stay competitive. However, upgrading fabrication capabilities at this stage requires billions more in equipment, labor, and calibration costs — at a time when order books are thin.
Trump’s Chip Push Meets Market Reality
The delay is awkwardly timed. President Donald Trump has made domestic semiconductor production a cornerstone of his second-term industrial policy, repeatedly emphasizing the need to reduce U.S. dependence on Asian suppliers. The CHIPS Act, passed with bipartisan support, aimed to reinvigorate local manufacturing by providing over $52 billion in funding and tax breaks to companies building in America.
But Samsung’s stalled progress underscores a fundamental miscalculation: Washington bet heavily on boosting production capacity but did not adequately anticipate that demand could cool amid global economic uncertainty, overcapacity, and restrictive U.S. policies on chip exports.
“Although yields have since improved, U.S. restrictions on high-end chip production for China have further weighed on the company, keeping its capacity utilization below the industry average,” Trendforce analyst Joanne Chiao said to Nikkei Asia.
While demand for chips in artificial intelligence and cloud computing remains strong, consumer electronics and automotive sectors — which account for a bulk of chip consumption — have pulled back. High interest rates, inflation, and trade friction have further constrained purchases, leaving fabs like Samsung’s Taylor plant in a holding pattern.
A growing number of industry leaders are now urging the U.S. government to reconsider its sweeping export controls, especially those targeting China. While intended to protect national security and limit China’s access to advanced semiconductors, the measures have also constrained revenue streams for U.S. chipmakers and their partners.
Nvidia CEO Jensen Huang has been one of the most vocal critics. Speaking at multiple forums, Huang warned that the U.S. restrictions — particularly the ban on selling high-performance AI chips to China — are creating unintended ripple effects.
Huang said Nvidia expected to lose up to $8 billion in sales in the second quarter alone due to its inability to supply chips to China—the world’s largest market for AI infrastructure.
“You cannot underestimate the importance of the China market,” he stressed. “This is the home of the world’s largest population of AI researchers.”
Huang emphasized that while safeguarding national security is important, a more nuanced, surgical approach to export controls is essential. Blanket bans, he argued, risk weakening U.S. firms while allowing Chinese competitors, who have enjoyed state backing to boost domestic production, to fill the void.
Samsung’s Taylor ordeal is a cautionary tale for U.S. industrial planners. While America has spent billions to lure chipmakers, it has not ensured that economic conditions and policy environments are stable enough to sustain those investments.
Unlike TSMC’s Arizona plant — which is fully booked through 2027 with orders from Apple, AMD, Broadcom, and Nvidia — Samsung has no major U.S. customers lined up. Its lower market share and struggles with advanced-node yields have further limited its appeal. TrendForce estimates that Samsung holds only 7.7% of the global foundry market, compared to TSMC’s dominant 68%.
In addition, building new fabs is only part of the equation. Companies must also construct robust local supply chains, hire thousands of skilled workers, and secure long-term clients — all while dealing with regulatory uncertainty and geopolitical friction.
While Samsung still says it plans to open the Taylor fab by 2026, no firm production date has been set. With billions already sunk into the facility and CHIPS Act funds contingent on actual operations, the company is under pressure to get the plant up and running — or risk being left behind.
Some analysts believe that unless U.S. policymakers adjust their strategy to better align production incentives with global demand and review export controls, more fabs could soon join Taylor in limbo.
