TSMC Smashes Records with Over 38% Profit Surge as AI Boom Fuels Demand

Taiwan Semiconductor Manufacturing Co. (TSMC), the world’s top provider of contract chip-making services, achieved record earnings in the third quarter, with profits rising by more than 38% from a year earlier. The surge surpassed market expectations and reflected strong, ongoing demand for advanced semiconductors used in artificial intelligence (AI) and 5G applications.

During the July–September period, TSMC reported net income of about T$452 billion (US$14.7 billion), outperforming analysts’ projections of roughly T$418 billion based on LSEG SmartEstimate data. Revenue reached approximately T$990 billion, up nearly 30% year over year, marking a second consecutive quarter of solid growth.

The company’s high-performance computing unit, which designs chips supporting AI and 5G workloads, represented close to 57% of total sales. In addition, semiconductors produced using 7-nanometer and smaller manufacturing processes contributed roughly three-quarters of wafer revenue, reinforcing TSMC’s lead in cutting-edge chip fabrication.

Industry observers noted accelerating demand for the firm’s most sophisticated chips, including those built on 3-nanometer and 4/5-nanometer technologies, with consistent orders from key customers such as Nvidia and Apple. This momentum illustrates how the global expansion of AI development continues to drive chip capacity utilization across TSMC’s most advanced plants.

Even amid uncertainty surrounding possible U.S. trade measures and tariff discussions, TSMC’s stock has climbed close to 40% so far this year, substantially outpacing the market’s roughly 20% rise. The company has also reaffirmed its US$100 billion investment plan to expand production capacity in the United States, a move aimed at strengthening supply-chain resilience and geographic diversification.

With six straight quarters of double-digit profit increases, TSMC’s latest financial performance underscores its pivotal role in powering the next generation of AI-centric semiconductor technologies.