TSMC, the world's largest semiconductor foundry, released its Q2 2026 financial results on Thursday (July 28), with revenue and net profit both surpassing market expectations, driven by the sustained surge in demand for AI chips. According to the data, TSMC's Q2 consolidated revenue was approximately NT$720 billion (about $24 billion), up 28% year-on-year and 12% quarter-on-quarter; net profit was approximately NT$280 billion, up 32% year-on-year, with earnings per share of NT$10.8, a record high for the same period.

AI Chip Orders Become Core Engine

TSMC management stated during the earnings call that the Q2 growth was mainly driven by strong demand for high-performance computing (HPC) and AI accelerator chips, with related revenue accounting for over 45% of total revenue for the first time. Among them, orders for the latest AI training and inference chips from customers such as Nvidia, AMD, and Broadcom pushed 3nm process capacity utilization to over 98%, essentially at full capacity.

"We are witnessing AI chips accelerating from the cloud to the edge. Whether it is large language models or autonomous driving, the demand for advanced processes is far from reaching its ceiling," said TSMC CFO Wendell Huang during the conference call. "We expect AI-related revenue in the second half of the year to grow by more than 30% compared to the first half."

Capex Raised to $42 Billion, Accelerating 2nm Mass Production

To meet strong customer demand, TSMC raised its full-year 2026 capital expenditure forecast from the original $40 billion to $42 billion, a record high for the company. About 70% will be used for capacity expansion and R&D of 3nm and 2nm advanced processes.

The company revealed that its 2nm process (N2) will be put into mass production ahead of schedule in early 2027, with initial customers including Apple, Nvidia, and Qualcomm. The current trial production yield has exceeded 80%, better than industry expectations. TSMC will introduce GAA (Gate-All-Around) transistor architecture for the first time at the N2 node, which is expected to improve energy efficiency by more than 15% compared to 3nm, providing key support for further cost reduction and efficiency improvement of AI chips.

Singapore Market Focus: Local Chip Companies Benefit Indirectly

Although TSMC does not have a direct factory in Singapore, its supply chain has boosted local semiconductor equipment and materials companies. Several semiconductor-related companies listed on the Singapore Mainboard, such as UMS Holdings and ST Engineering, saw slight gains in their stock prices today. Analysts believe that during the global AI chip expansion cycle, Singapore, as a hub for semiconductor packaging, testing, and materials supply in Southeast Asia, is expected to continue to benefit.

The head of research at Singapore brokerage Phillip Securities said: "TSMC's strong results confirm the high prosperity of the AI industry chain. For Singapore investors, focusing on TSMC ADR in their self-selected chip stock pool, as well as local supply chain targets such as UMS and AEM, will be a good choice."

Industry Outlook: AI Chip Demand Resilient Against Cyclical Fluctuations

Despite concerns from some market participants about a cyclical downturn in the chip industry, TSMC's earnings report sent a positive signal. The company expects revenue to continue to grow at a double-digit rate in the second half of the year, with full-year revenue growth expected to reach 25% to 30%.

"AI is the most certain growth engine for the semiconductor industry in the next decade," said TSMC Chairman Mark Liu at a media briefing. "We have seen AI applications expand from data centers to PCs, mobile phones, and automobiles, which will drive long-term demand for advanced processes."

It is worth noting that TSMC's ADR listed on the Singapore Exchange (SGX) has been actively traded recently, with average daily trading volume in Q2 increasing 40% compared to Q1, reflecting that more and more Singapore investors are choosing TSMC as the preferred target for allocating core AI assets.