The AI infrastructure dividend continues to be realized! TSMC's Q2 net profit surged 77% to a new high, raising the full-year capital expenditure and revenue guidance

According to Zhitong Finance APP, thanks to the strong global demand for artificial intelligence (AI) chips, Taiwan Semiconductor Manufacturing Company (TSMC) (TSM.US) announced a second-quarter profit that far exceeded expectations on Thursday. Data shows that TSMC's Q2 revenue was NT$1.27 trillion (US$39.45 billion), a year-on-year increase of 36%, with market expectations at NT$1.264 trillion; net profit was NT$706.562 billion, a year-on-year increase of 77.4% and a quarter-on-quarter increase of 23.4%, with market expectations at NT$632.64 billion.

The world's largest chip foundry has achieved record net profits for the fifth consecutive quarter. This performance is mainly attributed to the explosive growth in global AI infrastructure demand, which has driven strong orders for advanced process and advanced packaging technologies. TSMC stated that advanced process technologies of 7 nanometers and below accounted for 77% of total wafer revenue.

TSMC Raises Annual Capital Expenditure, Anchoring Long-term Demand for AI Chips

Notably, TSMC has raised its capital expenditure and full-year revenue expectations for this year, demonstrating the company's strong confidence in global AI chip demand.

TSMC currently expects its capital expenditure to reach US$60 billion to US$64 billion in 2026, up from the previous expectation of US$52 billion to US$56 billion. The company also anticipates that the revenue growth rate in US dollars will be slightly above 40%, higher than the previously predicted rate of over 30%.

The continued high growth in performance confirms TSMC's core position in the global AI industry chain: the vast majority of advanced chips used for data centers and smartphones are manufactured by it. This Asian company, with the highest market value, has become a key barometer for measuring the prosperity of global AI infrastructure.

On Thursday, TSMC confirmed that it may further increase spending over the next three years. This accelerated expansion reflects that large tech companies, from Meta to Alphabet, will continue to purchase large quantities of chips and hardware needed for building data centers.

TSMC's Chief Financial Officer Huang Wende told analysts during the earnings call: "We are very confident in the AI mega trend. Investment spending over the next three years will be much larger, far exceeding the past three years."

The additional spending may be allocated to Arizona, USA, where TSMC has reserved a total budget of US$265 billion for capacity expansion. TSMC President Wei Zhejia stated as early as June that even with a significant increase in domestic chip manufacturing capacity in the United States, the company would still be unable to meet the demands of American customers in the coming years.

Market Divergence Remains: The Expansion of Computing Power Carries Hidden Concerns

TSMC's capital expenditure is closely watched as a key leading indicator of global core chip supply and demand, covering categories from NVIDIA AI acceleration chips, Tesla in-vehicle processors, to core chips for servers that support various AI services.

South Korean chip giant SK Hynix also believes that the shortage of memory chips will continue until after 2030, as the spending spree by data center operators has stimulated demand for traditional memory and high-bandwidth memory (HBM) used in conjunction with AI systems.

This year alone, tech giants like Meta are expected to invest over US$725 billion in AI computing infrastructure. However, market divergence has emerged: the overall valuation of the AI sector is currently high, and many investors are concerned that the rush to expand computing power by major manufacturers may lead to an oversupply of chips In addition, leading data center operators are continuously taking on debt and financing to raise funds for the expansion of computing infrastructure, with a significant portion of their AI-related investments relying on the steadily rising debt. Despite the accelerating pace of the global AI industry, there remains great uncertainty about whether the massive investments can yield substantial returns.

Since the beginning of this year, driven by the explosive demand for global AI computing power, the global semiconductor sector has seen a significant rise. As the earnings season approaches, the market has begun to reassess the valuation of the chip industry, and whether the AI market can sustain itself has become a focal point. On Thursday, Asian chip stocks faced collective pressure, dragging down the overall market index, which fully reflects current investors' concerns about the high valuations in the AI industry chain