As demand for artificial intelligence (AI) chips increases, a survey found that revenue in the global "foundry 2.0" market in the first quarter of 2026 rose 23% from a year earlier to $86 billion (about 133.18 trillion won). Demand for AI graphics processing units (GPUs) and application-specific integrated circuits (ASICs) lifted both advanced-node wafer shipments and packaging utilization.
Market research firm Counterpoint Research on the 30th released this analysis through its foundry market supply tracker. Foundry 2.0 is a concept that includes not only pure-play foundries but also non-memory integrated device manufacturers (IDMs), outsourced semiconductor assembly and test (OSAT) corporations, and photomask suppliers.
By field, pure-play foundries accounted for 57% of the total market. Non-memory IDMs made up 29%, OSAT 13%, and photomasks and other segments 1%. In the first quarter of 2026, revenue in the pure-play foundry institutional sector increased 31% from a year earlier, while OSAT and non-memory IDMs rose 21% and 12%, respectively.
TSMC ranked first with a 38% share of the overall foundry 2.0 market. Samsung Foundry was tallied at 4%. SMIC recorded 3%, and UMC 2%. In the non-memory IDM institutional sector, Intel Foundry and Texas Instruments each held 6%, and Infineon Technologies AG stood at 5%.
TSMC was cited as the biggest beneficiary corporations of the AI Semiconductor upcycle. The company's first-quarter 2026 revenue increased 41% from a year earlier. Demand for AI GPUs, AI ASICs, and advanced packaging supported high utilization at advanced nodes. Counterpoint projected that if this trend continues, TSMC's revenue growth this year will reach about 36%.
William Li, a research director at Counterpoint, said, "What stands out in this cycle is not only the strength of AI demand, but also that this trend is changing TSMC's operating strategy itself," adding, "An unprecedented level of capacity reallocation is underway across multiple production lines."
Growth by Chinese players also continued in the pure-play foundry market. SMIC's revenue in the first quarter of 2026 rose 12% from a year earlier, and Nexchip increased 19%. Demand for semiconductor localization in China and higher wafer prices helped improve results.
UMC and Vanguard, which focus on mature nodes, also increased revenue. The two companies' first-quarter revenue rose 10% and 14% year over year, respectively. A recovery in consumer electronics demand and strength in the power management IC (PMIC) market are cited as drivers. Counterpoint analyzed that as TSMC concentrates on advanced nodes, these companies could absorb volumes that TSMC cannot accommodate.
Demand for AI Semiconductor is also spreading to back-end processes. The OSAT institutional sector grew 21% in the first quarter of 2026 from a year earlier. ASE's revenue increased 18%, and Amkor rose 25% on the back of high utilization at advanced packaging lines. Tongfu saw revenue climb 29% on the expansion of AMD's AI packaging, while KYEC grew 45% amid lengthening AI test lead times.
Brady Wang, a research fellow at Counterpoint, said, "Advanced packaging is emerging as a key supply constraint in the spread of AI," noting, "OSAT companies are securing both a stable demand base and profitability on the back of solid first-quarter 2026 results and aggressive capacity expansion."
Counterpoint saw the trend of diversifying supply sources as a potential opportunity for Intel Foundry and Samsung Foundry. As AI demand expands, shortages in advanced-node and packaging capacity are growing, prompting major customers to move to reduce reliance on a single supply chain.
Intel is pushing the introduction of advanced packaging technology and the expansion of its foundry business. Counterpoint cited the possibility that Apple's next-generation M-series chips could be awarded to the Intel 18A-P process as a factor strengthening Intel Foundry's competitiveness. Samsung Foundry was also mentioned as a candidate to gain new orders as AI Semiconductor customers diversify their supply chains.
Counterpoint explained that while the existing foundry 1.0 centers on chip manufacturing, foundry 2.0 reflects an integrated ecosystem in which design, manufacturing, packaging, and testing are connected. The analysis is that as AI systems become more complex, not only node scaling but also advanced packaging and testing supply capabilities are emerging as core competitiveness in the semiconductor industry.