Shares
TSMC & ASML Q2 2025 earnings: AI chips demand meets tariffs
The latest quarterly earnings from key semiconductor players underscore the powerful tailwinds from the AI boom — and the growing risks posed by Trump’s tariffs, which threaten global economic stability and fragile chip supply chains. ASML posted strong Q2 2025 results but warned that sales growth may stall next year. Meanwhile, TSMC struck a more optimistic tone, citing sustained demand for AI chips. With Nvidia expected to resume H20 chip sales to China and Big Tech firms like Meta and Google ramping up AI investment, the semiconductor market remains well-positioned for continued growth.

Trump tariffs threaten semiconductor supply chain
Donald Trump’s sweeping tariffs could trigger a global economic slowdown, with the World Bank recently cutting its 2025 GDP forecast by 0.4% [1]. It cited “increased trade tension and heightened policy uncertainty” as key reasons for the expected deceleration in major economies such as the US and China. Macroeconomic headwinds, US export restrictions on advanced technologies, and a 50% tariff on copper imports all threaten the intricate supply chains of the semiconductor industry. These factors may constrain investment in data centres designed to support artificial intelligence (AI), thereby weakening demand for semiconductors.
Moreover, economic uncertainty can depress as sales – a core income source for tech giants like Meta and Google – potentially hindering their AI-related spending. Meta’s CFO, Susan Li, highlighted advertising risks in the most recent earnings call, noting “some reduced spend” from Asia-based e-commerce companies [2]. Alphabet’s Chief Business Officer, Philipp Schindler, similarly warned of a “slight headwind” for the firm’s ad business this year. [3]
Cracks in the advertising market are also evident in the latest updates from WPP and Omnicom – both members of the so-called Big Six marketing groups. WPP downgraded its forecasts and now expects sales to decline by 3–5% this financial year, citing “continued macro uncertainty weighing on client spend” [4]. Omnicom maintained its full-year guidance of 2.5–4.5% growth, although this is below its 2024 performance. The firm also flagged risks from “ongoing macroeconomic and geopolitical uncertainty”. [5]
AI boom fuels global semiconductor demand
DeepSeek’s breakthrough earlier this year marked a turning point in the rapid expansion of AI, sparking increased efforts by Chinese companies such as Alibaba and Baidu. This has challenged US dominance in the sector and prompted major American tech firms to double down on their AI investments.
Meta Platforms is among the frontrunners in this race for AI leadership, expecting at least a 63% year-on-year increase in capital expenditures this year. Earlier this month, CEO Mark Zuckerberg announced plans to invest hundreds of billions of dollars to develop Superintelligence [6] . Alphabet, Google’s parent company, also anticipates elevated spending.
Even as the cost of developing large language models declines, the proliferation of AI is expected to multiply inference workloads – thereby driving further demand for semiconductors and data centres. Following Nvidia’s most recent earnings, CEO Jensen Huang spoke on Bloomberg of “off the charts” demand for reasoning-based AI inference as services like ChatGPT grow in popularity. The World Semiconductor Trade Statistics (WSTS) projects an 11.2% increase in global chip sales this year [7], while the International Energy Agency (IEA) forecasts a 133% rise in data centre capacity by 2030, reaching 226 gigawatts. [8]
Nvidia is the go-to solution for AI training and inference solution, with the company citing “incredibly strong” demand for its Ai infrastructure [9]. Although its impressive revenue growth is slowing, an anticipated resumption of H20 chip sales in China could renew momentum [10]. These chips are less powerful to comply with US export restrictions. However, sales were previously banned, and Nvidia expects this disruption to cost it $8 billion in lost revenue this quarter.
TSMC & ASML earnings highlight AI growth and trade risks
Taiwan Semiconductor Manufacturing Company (TSMC), which produces advanced chips used in AI systems designed by Nvidia and others, reported a 38.6% year-on-year increase in second-quarter revenue, according to Thursday’s earnings [11]. This demonstrates resilience despite broader macroeconomic challenges. Notably, profits surged 60.7% – outpacing revenue –, showcasing the importance of these more expensive advanced AI chips that offer higher margins.
To manufacture these chips, TSMC relies on equipment from ASML – another critical chokepoint in the global semiconductor supply chain. Benefiting from strong AI-driven demand, the Dutch company posted robust second-quarter results on Wednesday, beating estimates. Revenue rose 23.2% to €7.7 billion, bookings remained strong, and margins held firm at 53.7%. [12]
However, ASML narrowed its 2025 forecast, now expecting just a 15% rise in sales. Crucially, it was unable to confirm growth for 2026, citing “increasing uncertainty driven by macroeconomic and geopolitical developments”.
AI drives chips growth in 2025 - but trade uncertainty looms
The global semiconductor industry is set for continued growth this year as the AI revolution fuels demand. Major tech firms are investing heavily to expand their capabilities and launch new AI services, leading to more data centres and higher chip consumption. Recent earnings from TSMC and ASML underscore this robust demand, resulting in significant revenues and profits.
However, trade barriers and export restrictions remain significant threats. They risk disrupting the intricate semiconductor supply chain, limiting investment and reducing demand. Despite the sector's current resilience, cracks are beginning to show – and ASML’s warning highlights the mounting risks ahead.

Senior Financial Editorial Writer
Nikos Tzabouras
Nikos Tzabouras is a graduate of the Department of International & European Economic Studies at the Athens University of Economics and Business. With extensive experience in market analysis and a strong foundation in international relations, he brings a unique perspective to financial markets. Nikos emphasizes not only technical analysis but also on fundamentals and the growing influence of geopolitics on financial trends.
As a Senior Financial Editorial Writer, he delivers comprehensive and forward-looking insights across a wide range of asset classes, including equities, commodities, and currencies. His work explores how macroeconomic events, political developments, and global policies impact market dynamics, providing readers with a deeper understanding of both short-term movements and long-term trends.