According to the latest World Fab Forecast report by SEMI, the International Semiconductor Industry Association, the rising demand for chips is driving continuous growth in global semiconductor wafer fab capacity, with increases of 6% and 7% in 2024 and 2025, respectively, reaching a record monthly capacity of 33.7 million 8-inch equivalent wafers.
SEMI states that with data center training and inference for processes below 5 nanometers, and the advancement of generative AI technologies, there is expected to be a 13% growth in 2024. To improve chip efficiency, major manufacturers like Intel, Samsung, and TSMC are preparing to produce 2-nanometer gate-all-around (GAA) chips in 2025, leading to a 17% increase in total advanced process capacity.
SEMI’s global marketing director and president of SEMI Taiwan, Terry Tsao, analyzed that AI’s pervasive presence from cloud computing to various edge devices intensifies the competition for high-performance chip development, driving strong expansion in global semiconductor manufacturing capacity and creating a positive feedback loop: AI accelerates the growth of various semiconductor applications, stimulating higher investment.
Chinese chip manufacturers are expected to maintain double-digit capacity growth, with a 15% increase this year to 8.85 million wafers per month, and a further 14% growth in 2025 to 10.1 million wafers per month, nearly one-third of the industry’s total volume.
Other major chip manufacturing regions are projected to have capacity growth of no more than 5% by 2025. Taiwan ranks second with a monthly output of 5.8 million wafers (a 4% increase). South Korea is expected to grow by 7% in 2025, reaching 5.4 million wafers per month after surpassing 5 million wafers for the first time this year, ranking third. Semiconductor capacities in Japan, the Americas, Europe, the Middle East, and Southeast Asia are projected to be 4.7 million wafers per month (up 3% annually), 3.2 million wafers (up 5% annually), 2.7 million wafers (up 4% annually), and 1.8 million wafers (up 4% annually), respectively.
Benefiting from Intel’s establishment of wafer foundry services and China’s capacity expansion, the foundry sector’s capacity will grow by 11% in 2024 and another 10% in 2025, reaching a monthly output of 12.7 million wafers by 2026.
The rapid growth in AI server computing power is driving the demand for HBM (High Bandwidth Memory), providing long-unseen growth momentum for the memory sector. The explosive deployment of AI requires tightly stacked HBM, with each HBM integrating 8-12 dies, which is why many DRAM market leaders are increasing investment in HBM/DRAM. DRAM capacity is expected to grow by 9% this year and next, while the recovery in the 3D NAND market is slower, with no capacity increase this year and a 5% rise in 2025.
The rise of AI applications in edge devices is increasing the DRAM capacity of mainstream smartphones from 8GB to 12GB, and laptops equipped with AI assistants require at least 16GB of DRAM. The expansion of AI to edge devices is expected to further drive up DRAM demand.
Related:
- Key Differences in Three Silicon Wafer Orientations
- Understanding Chamfering in Semiconductor Processes
- Best Way to Clean Silicon Wafers for Top Performance
- Here’s Why Wafer Carriers Always Hold 25 Wafers Only
- Master ALD Thin Film Deposition in One Quick Guide
- US Value Clause Reshapes 2025 Device Shipments
- Apple Moves from 3nm to 2nm in Upcoming iPhone 18
- Global Wafer Fab Equipment Revenue Hits $133 Billion in 2024
- Apple AI Tech Boom, Market Value Nears $4 Trillion
Disclaimer: This article is created by the original author. The content of the article represents their personal opinions. Our reposting is for sharing and discussion purposes only and does not imply our endorsement or agreement. If you have any objections, please get in touch with us through the provided channels.