Taiwan Semiconductor Manufacturing (TSM.US), the world's leading chip foundry and a key supplier to Nvidia and Apple, reported a staggering 58% surge in first-quarter net profit. This latest data from the AI compute supply chain indicates that recent geopolitical tensions in the Middle East have not dampened booming global investment in AI compute infrastructure or the explosive worldwide demand for AI processing power. For the quarter ending in March, TSMC achieved a net profit of NT$5.725 trillion (approximately $180 billion), significantly surpassing the average analyst forecast of NT$5.424 trillion—an estimate that had been consistently revised upward since February.
Regarding forward-looking guidance, which is a key market focus, TSMC's management provided an outlook that comprehensively exceeded recently upgraded market expectations. The company now anticipates total revenue growth for the year to exceed 30%, up from a previous expectation of below 30% and above the consensus forecast of 25%-28%. Company executives also stated that, given persistently strong demand for AI chips and advanced packaging, capital expenditures are expected to be near the upper end of the previously guided range (up to $56 billion). They emphasized that capital spending over the next three years will be significantly higher than in the past three years, signaling strong confidence in future growth prospects. TSMC projects second-quarter revenue between $39 billion and $40.2 billion, well above market expectations of $38.1 billion. It also forecasts a gross margin of 65.5% to 67.5% for the second quarter, exceeding the market's estimate of 64%.
It is noteworthy that these performance expectations have been repeatedly upgraded since late January, following robust earnings reports from U.S. tech giants like Google, Amazon, and Nvidia. Even against this backdrop, TSMC's official outlook remains stronger than the analysts' already elevated expectations, underscoring the sheer intensity of global demand for AI compute infrastructure—driven by AI GPU/ASIC technologies—and the surge in requirements for high-performance networking chips and customized storage controllers linked to AI data center construction. TSMC's growth trajectory continues to follow the explosive pattern seen since 2024, reminiscent of Nvidia's performance.
TSMC's first-quarter revenue increased by 35% year-over-year, also significantly beating market forecasts. Total Q1 revenue reached NT$11.34 trillion, up 35% year-over-year, exceeding the projected NT$11.2 trillion. Operating profit was NT$659 billion, also surpassing the average analyst estimate of NT$623.8 billion.
From a process technology perspective, advanced nodes remain TSMC's core revenue driver. The 5nm process contributed the largest share at 36%, followed by the 3nm process at 25%. Combined, processes at 5nm and below accounted for over 60% of total revenue, showing a trend of accelerating growth quarter-over-quarter and highlighting the increasing preference for the most advanced 3nm and below processes in AI chip manufacturing.
The gross margin was a standout highlight of TSMC's first-quarter performance. TSMC's Q1 gross margin rose to a record 66.2%, expanding nearly 4 percentage points from the previous quarter's 62.3% and substantially exceeding the market's average expectation of 64.5%. The operating profit margin surged to 58.1%, above the market's forecast of 55.6% and the previous quarter's 54.0%, demonstrating continued expansion of the profit structure underpinned by explosive AI compute demand.
TSMC's strong results will likely help alleviate market concerns that the Middle East geopolitical crisis, and its impact on oil prices, could weaken demand for high-cost, high-power AI data centers or even consumer electronics like iPhones. The conflict has already pressured global shipping routes and energy costs, and investors are seeking clues about whether its effects might spread to tech giants' unprecedented trillion-dollar AI compute spending plans. TSMC's exceptionally robust performance clearly illustrates that as model scale, inference chains, and multimodal/agentic AI workloads drive exponential growth in compute resource consumption, tech giants are increasingly concentrating their capital expenditures on AI compute infrastructure. Global investors continue to anchor the "semiconductor bull market narrative"—centered on Nvidia, Google's TPU clusters, AMD's new product cycles, and AI cluster delivery expectations—as one of the most certain growth stories in global equity markets. This also implies that investment themes closely tied to AI training and inference, such as power supply, liquid cooling systems, and optical interconnect supply chains, will remain among the stock market's hottest sectors, led by Nvidia, AMD, Broadcom, TSMC, and Micron, even amid Middle East geopolitical uncertainties.
Equity strategists from BlackRock, the world's largest asset manager, have shifted back to an "overweight" stance on U.S. and emerging market equities, primarily because they believe the substantive damage to global economic growth from the new Middle East conflict is "likely to be very manageable." Regarding core investment themes, BlackRock strategists are particularly bullish on semiconductor stocks closely linked to AI compute infrastructure, including leaders in the U.S., South Korean, and Taiwan stock markets.
As illustrated, TSMC's stock price has hit record highs driven by the AI boom—Nvidia and most other AI chip giants are heavily reliant on TSMC's manufacturing capabilities. TSMC's strong rally and market cap expansion have helped the Taiwan stock market's total valuation surpass that of the U.K. for the first time.
Although Google parent Alphabet, Amazon, Facebook parent Meta Platforms, Microsoft, and Oracle have collectively pledged to allocate at least approximately $6.5 trillion for AI compute infrastructure spending this year, some Wall Street analysts have questioned whether TSMC and its major clients like Nvidia can sustain or even expand their current robust growth rates. Concerns have been raised about potential spending cuts if these tech giants face cash flow constraints or if actual cloud AI compute demand proves less optimistic than previously expected. However, TSMC's latest strong results and future outlook have effectively dismantled these doubts regarding global AI data center construction and AI compute demand.
Following explosive sales that propelled Nvidia to become the world's most valuable company and TSMC to become Asia's largest company by market cap, investors are seeking strong assurances that the booming AI compute spending can continue. Driven by TSMC's strong stock performance this year, the Taiwan stock market's total valuation has risen to $4.14 trillion, surpassing the U.K. to rank seventh globally. Tech companies like TSMC, Foxconn, and Wistron, which are closely tied to the unprecedented AI data center construction wave, continue to attract global capital inflows.
The market also speculates that a prolonged Middle East crisis could disrupt the supply of key chip manufacturing components and gases like helium, whose production is concentrated in Middle Eastern countries. Equipment supply constraints could also limit growth in the potentially $1 trillion chip industry chain, as semiconductor equipment companies like ASML Holding NV, Applied Materials, and Lam Research may struggle to expand capacity quickly enough to meet demand for cutting-edge equipment from clients like TSMC.
Addressing concerns about supply chain impacts from the Middle East conflict, TSMC management reassured investors during the earnings call, stating that the company is sourcing specialty gases from multiple suppliers in different regions and does not currently anticipate material supply disruptions. TSMC also noted that prices for certain chemicals and gases might rise due to the conflict, but it is too early to quantify the impact on TSMC, adding that helium and hydrogen inventories remain sufficient.
TSMC manufactures the vast majority of the world's most advanced chips, particularly holding nearly all capacity for 5nm and below processes, as well as advanced packaging like CoWoS/3D. As the primary foundry partner for Nvidia, AMD, Broadcom, and Apple, TSMC has become a central beneficiary of the global AI boom amid the race to build AI compute infrastructure clusters. Year-to-date, TSMC's U.S.-traded ADRs have risen about 25%, significantly outperforming major clients like Nvidia.
Concurrently, the company is addressing another concern: a persistent shortage of memory chips—a component TSMC does not primarily produce. However, a severe and ongoing memory chip shortage could lead to the first significant contraction in the global smartphone market since 2023. The chipmaker indicated earlier in 2026 that it primarily supplies the global high-end AI smartphone market, where demand remains robust and less sensitive to memory component price increases.
As chips become core strategic assets for nations and tech giants competing in the unprecedented AI wave, TSMC faces competitive pressure from longstanding rivals Intel and Samsung in advanced 3nm and below process foundry services, as well as from new entrants attempting to join this lucrative field. For example, Tesla CEO Elon Musk is advancing his ambitious Terafab large-scale chip manufacturing project to produce chips for his companies' substantial AI compute infrastructure needs, including Tesla, SpaceX, and xAI. Meanwhile, the Japanese government is heavily funding Rapidus Corp., aiming for this Japanese AI startup to begin production of cutting-edge 2nm and below chips by 2027.
The Middle East conflict appears unable to halt the AI frenzy! The global AI arms race, led by Google, Meta, and Microsoft, is far from over. TSMC's first-quarter results—a 58% year-over-year surge in net profit, 35% revenue growth exceeding expectations, and an outlook that crushes all forecasts—send a very clear positive signal to the entire AI compute supply chain: AI demand has not been extinguished by high oil prices, raw material shipping disruptions, or energy volatility caused by the Middle East war. Instead, it continues to push advanced process AI chips and CoWoS advanced packaging to full capacity.
Furthermore, stocks directly linked to AI compute infrastructure—tagged with "earnings certainty + high beta attributes"—and led by the "AI compute super group" of Nvidia, TSMC, AMD, and Broadcom, typically represent the most sensitive, first-moving, and highest-gaining layer within broader market or tech stock rebound logic. The core rationale behind this is exceptionally solid: this segment is directly tied to tech giants' record-breaking AI capital expenditures, not merely speculative narratives.
AI hyperscalers (super cloud computing giants like Google, Microsoft, and Amazon) are continuing their capital expenditure arms race. As long as they "prefer to take on debt and implement layoffs rather than retreat in the AI capex race" (the so-called "AI compute arms race"), the leaders across the AI compute supply chain remain attractive for investment.
For global equity markets and AI compute supply chain allocation, the most critical signal from these earnings is that the AI compute super-cycle has not been disproven in any way. Instead, it is further converging towards a few truly beneficial, hardcore supply-demand bottleneck areas within the core technology chain. The most certain future beneficiaries remain those companies leading the AI compute supply chain—those controlling the supply of the world's most advanced AI chips, advanced packaging, HBM/eSSD/enterprise-grade DRAM storage, enterprise server CPUs, data center optical interconnect equipment, AI power chains, and key semiconductor equipment.
A recent forecast from Bank of America strategists projects that the global semiconductor market will reach a total size of $2 trillion by 2030, driven by accelerated growth in core AI compute supply chain leaders (led by Nvidia, Broadcom, TSMC, and Marvell Technology, with forward P/E ratios between 15x-20x) and in areas like memory/logic chips, 2.5D/3D advanced packaging, and data center power chains. The projected compound annual growth rate from 2025 to 2030 is 20%. In contrast, the global semiconductor market size is expected to be below $1 trillion until at least 2025.
According to the latest compiled analyst estimates, Amazon, along with Google parent Alphabet, Facebook parent Meta Platforms, Oracle, and Microsoft, are projected to accumulate approximately $6.5 trillion in AI-related capital expenditures by 2026. Some analysts believe total spending could exceed $7 trillion—implying a potential year-over-year increase in AI capital spending of over 70%. Notably, these five U.S. super tech giants are expected to invest a cumulative total of about $15 trillion between 2023 and 2026 to build massive AI compute infrastructure. This compares to a total investment of approximately $6 trillion by these companies throughout their entire history prior to 2022.