AI Keeps Chips Hot

The chip industry’s latest earnings say the artificial intelligence buildout is still running hot, with Taiwan Semiconductor Manufacturing Co. at the center. (investor.tsmc.com) Taiwan Semiconductor Manufacturing Co., the world’s largest contract chipmaker, reported first-quarter 2026 revenue of $35.9 billion, gross margin of 66.2%, and operating margin of 58.1% on April 16. Net profit rose 58% from a year earlier to about T$572.5 billion, or roughly $18.2 billion. (investor.tsmc.com; reuters.com) The company guided second-quarter revenue to $39.0 billion to $40.2 billion and said full-year 2026 revenue should grow by more than 30% in U.S. dollar terms. Reuters reported that customers including Nvidia and Apple helped drive the quarter as demand for artificial intelligence processors stayed strong. (investor.tsmc.com; reuters.com) The numbers fit a broader pattern across the supply chain. ASML, the Dutch maker of the lithography tools used to print advanced chips, reported €8.8 billion in first-quarter 2026 net sales on April 15 and lifted its 2026 outlook. (asml.com; finance.yahoo.com) SK hynix, one of the main suppliers of high-bandwidth memory used alongside artificial intelligence processors, reported first-quarter revenue of 52.5763 trillion won and operating profit of 37.6103 trillion won on April 23. The company said strong demand persisted despite a normal seasonal slowdown and pointed to higher sales of HBM, server DRAM, and enterprise solid-state drives. (news.skhynix.com) At the demand end, Nvidia reported $68.1 billion in quarterly revenue for the period ended January 25, 2026, up 73% from a year earlier, with data center revenue of $62.3 billion, up 75%. That is the spending wave now flowing backward through foundries, memory makers, and equipment suppliers. (nvidianews.nvidia.com) A chip foundry is a factory that manufactures designs from companies like Nvidia, and a lithography company sells the machines that etch those designs onto silicon. When all three layers — designer, factory, and toolmaker — post strong quarters at the same time, it points to capacity being added rather than only prices rising. (nvidianews.nvidia.com; investor.tsmc.com; asml.com) That does not mean supply constraints are gone. Reuters reported on April 23 that SK hynix said artificial intelligence memory demand would exceed manufacturing capacity, while CNBC reported on April 15 that ASML still faces tighter China export restrictions even after raising guidance. (reuters.com; cnbc.com) The latest earnings do show that more compute is coming online in 2026 than many buyers could get in 2024. As chip supply broadens, the bottlenecks for artificial intelligence use shift more toward how systems are designed, what data they can access, and what rules companies and governments impose on them. (investor.tsmc.com; news.skhynix.com; asml.com) For now, the cleanest signal is still the simplest one: the companies that make chips, memory, and chipmaking tools are all reporting that artificial intelligence customers keep ordering more. (investor.tsmc.com; asml.com; news.skhynix.com; nvidianews.nvidia.com)