Taiwan Semiconductor Manufacturing Company (TSMC) faces increasing demand for advanced chips, creating capacity constraints that are beginning to impact partner firms.
Illustration by IMF Alpha editorial · Reviewed by IMF Alpharoom AI
Taiwan Semiconductor Manufacturing Company (TSMC) is grappling with significant capacity constraints, particularly in its advanced process technologies, as global demand for artificial intelligence (AI) semiconductors escalates. The company, a foundational supplier for many major chip designers, reported a 16.5% year-over-year revenue increase in Q1 2024, reaching $18.87 billion, driven largely by high-performance computing (HPC) and AI applications.
Industry analysts indicate that lead times for some of TSMC's cutting-edge nodes, such as the 3-nanometer (N3) and 5-nanometer (N5) processes, have extended to several months. This situation is particularly challenging for companies like AMD and Broadcom (AVGO), which rely heavily on TSMC for manufacturing their latest AI accelerators and data center chips. While TSMC has announced plans to expand capacity, new fabs require substantial capital expenditure and several years to become fully operational.
ASML Holding NV (ASML), a critical supplier of extreme ultraviolet (EUV) lithography equipment to TSMC, also faces pressure to increase output. ASML's Q1 2024 net sales were €5.29 billion, down from €6.75 billion in Q4 2023, partly due to the timing of system shipments. However, the company projects strong demand growth in the latter half of 2024 and into 2025 as chipmakers like TSMC accelerate equipment installations.
The global semiconductor supply chain remains tight, with the ripple effects of TSMC's capacity limitations extending to end-product manufacturers. Companies anticipating new AI hardware launches may experience delays or higher production costs due to the competitive landscape for foundry services. This environment underscores the strategic importance of geographical diversification in chip manufacturing, a trend supported by initiatives such as the CHIPS Act in the United States.
TSMC has allocated a significant portion of its capital expenditure for 2024, projected between $28 billion and $32 billion, towards N3 and N2 process technologies development and expansion. Despite these investments, the full resolution of current capacity constraints is not expected in the immediate term, suggesting that the tight supply environment for advanced chips will persist through at least 2025.
Recent fintech earnings reports highlight varied payment volume growth and the increasing integration of AI in credit underwriting processes by major players.
As privacy rules and model hunger collide, synthetic data marketplaces are exploding — but investors and engineers should watch the realism gap and provenance problem.
How local large language models are reshaping privacy, app economics, and the chip wars—what consumers and investors need to know now.