ASML and TSMC Q2 Results Reveal Distinct Monopolies
Both semiconductor giants beat Q2 2026 expectations, but their earnings drivers diverge sharply between equipment upgrades and packaging bottlenecks.
ASML and TSMC both exceeded Q2 2026 revenue expectations, yet their underlying business mechanics remain distinct. ASML reported $10.65 billion in revenue, a 21.3% year-over-year increase, driven largely by its installed base services. TSMC generated $40.20 billion in revenue, up 36.0% annually, with high-performance computing accounting for 66% of total sales. These figures highlight different positions within the AI infrastructure supply chain rather than simple cyclical exposure.
The market often mischaracterizes ASML as a purely cyclical tool vendor and TSMC as a standard foundry. However, ASML’s $45.06 billion backlog and TSMC’s packaging constraints suggest deeper structural moats. ASML’s Installed Base Management segment contributed $3.15 billion, indicating steady demand for upgrades on existing tools. TSMC’s CEO C.C. Wei noted that packaging capacity is currently a limiting factor for customer growth, positioning the company as a bottleneck operator rather than just a manufacturer.
Service Revenue Drives ASML Growth
ASML’s Q2 performance was anchored by its service and upgrade business, which generated $3.15 billion. This segment provides stability distinct from lumpy equipment sales. CEO Christophe Fouquet attributed demand to ongoing AI investments in advanced logic and memory chips. The company’s gross margin stood at 54.0%, reflecting the high value of its installed base management offerings. This recurring revenue stream contrasts with the perception of ASML as solely dependent on new equipment cycles.
Investors often focus on EUV export restrictions, overlooking ASML’s near-monopoly in immersion DUV systems. The company plans to ship approximately 130 immersion DUV systems this year and expand capacity by 30% for 2027. China remains a significant market, representing about 20% of full-year sales, driven by mainstream node demand. This broad product portfolio ensures that ASML’s revenue is not solely tied to the most advanced lithography tools, providing a buffer against geopolitical trade shifts.
TSMC Packaging Capacity Limits Output
TSMC’s Q2 gross margin reached 67.7%, supported by strong high-performance computing demand, which grew 20% quarter-over-quarter. The company’s advanced packaging, specifically CoWoS, is a critical bottleneck for AI accelerators from Nvidia and AMD. C.C. Wei stated that packaging capacity is so tight it limits customer growth. This scarcity allows TSMC to maintain pricing power, although the CEO rejected the idea of extreme price hikes, citing a preference for stable pricing over short-term gains.
The concentration of TSMC’s revenue in HPC highlights its central role in the AI buildout. With 66% of revenue from this segment, TSMC is less diversified than its peers, making it highly sensitive to AI infrastructure spending. The company’s ability to scale CoWoS production is therefore a key determinant of its future earnings. This strategic position differentiates TSMC from other foundries that lack comparable advanced packaging capabilities, reinforcing its dominant market share in leading-edge nodes.
Guidance Highlights Margin and Capacity Risks
ASML guided Q3 revenue to €11.0–€12.0 billion and raised its full-year 2026 outlook to €43–€45 billion. This upward revision reflects confidence in its multi-year visibility, particularly from the 2028 investigation into future capacity. The company’s forward P/E is near 28, a premium that assumes the sustained demand for its lithography tools will materialize. Investors are watching for proof that ASML can convert its long-term investigations into firm orders, which would validate its scarcity premium.
TSMC guided Q3 revenue to $44.6–$45.8 billion but flagged a 3 to 4 point margin dilution from the 2nm node ramp. This margin compression is a key data point to monitor, as it could impact the company’s high profitability. TSMC’s forward P/E is near 20, reflecting its scale and operating leverage. The critical question is whether the company can close the packaging capacity gap it described as significant. If margins compress further, it may signal that the AI buildout is slowing or that costs are rising faster than revenue.
