Demand for AI computing power explodes! Samsung raises chip foundry prices: up to 15%

Kuai Technology reported on August 19 that, driven by the continued outbreak of global demand for AI computing power chips, the supply of advanced process foundry production capacity has become tight. Samsung Electronics has increased its foundry quotations for new orders, with the highest increase for some process nodes reaching 15%. This price adjustment has been officially implemented since July and is only for new orders. Existing long-term cooperation orders that have been signed and completed will not be affected by this price adjustment. In terms of specific process nodes, the OEM prices for the 4nm SF4 process have increased by 10%-15%, the 5nm SF5 process has increased simultaneously by 10%-15%, and the OEM quotations for the 8nm mature process have increased by nearly 10%. Samsung's 4nm production line in Pyeongtaek, South Korea, has been operating at full capacity since the end of last year. The production line not only accepts logic chip orders from external customers, but also produces its own HBM high-bandwidth memory supporting base chips, so production capacity resources are very tight. At the industry level, TSMC's advanced process production capacity continues to be saturated, and a large number of AI and high-performance computing orders are spilling out, giving Samsung's foundry business more customers and bargaining space. Previously, Samsung's wafer foundry sector has been in a state of loss for a long time, and this round of price increases is regarded as an important signal for business improvement. According to Samsung's internal forecast, advanced process revenue will account for more than 50% of total foundry revenue in 2026, and AI and high-performance computing-related businesses will account for more than 30%, achieving significant growth compared with the 15%-20% proportion at the end of 2025. The simultaneous price increases of leading wafer fabs will also be transmitted downstream. The manufacturing costs of AI acceleration chips, mobile phone SoCs, and automotive chips will rise accordingly, and terminal equipment manufacturers may face upward pressure on component costs.