Fab Capacity Utilization — Historical Time Series Data
TSMC, UMC, and Samsung foundry utilization rates quarterly from Q1 2022 through Q1 2026. Shows the two-speed recovery: leading-edge at 100% vs mature nodes declining. Samsung foundry series tracks the yield crisis and customer defections to TSMC.
Fab Capacity Utilization
TSMC, UMC, and Samsung foundry utilization rates quarterly from Q1 2022 through Q1 2026. Shows the two-speed recovery: leading-edge at 100% vs mature nodes declining. Samsung foundry series tracks the yield crisis and customer defections to TSMC.
Data sparse before 2022. Public pricing data for this category is limited prior to the AI demand inflection. Data points before 2022 are single-source estimates and should be treated as directional only.
5 data point values hidden. Unlock exact values, sources, and exports with Pro.
Data Table
| Period | Series | Value |
|---|---|---|
Q3 2022 | TSMC Leading Edge (≤7nm) | ••• |
Q3 2022 | UMC Mature (>14nm) | ••• |
Q4 2022 | TSMC Leading Edge (≤7nm) | ••• |
Q1 2023 | UMC Mature (>14nm) | ••• |
Q1 2023 | TSMC Leading Edge (≤7nm) | ••• |
Q2 2023 | TSMC Leading Edge (≤7nm) | 90% |
Q3 2023 | UMC Mature (>14nm) | 75% |
H2 2023 | TSMC Leading Edge (≤7nm) | 100% |
2024 | UMC Mature (>14nm) | 65% |
2024 | TSMC Leading Edge (≤7nm) | 100% |
2024 (TSMC mature) | TSMC Mature (≥14nm) | 70% |
H1 2025 | UMC Mature (>14nm) | 68% |
H1 2025 (TSMC mature) | TSMC Mature (≥14nm) | 72% |
Methodology & Sources
Sources: TSMC quarterly earnings transcripts, SemiAnalysis earnings analysis, Deep Research (120 sources). Note: foundries do not disclose exact utilization by node; figures are analyst estimates from wafer shipment and revenue data.
Citations:
- SemiAnalysis (Jan 2023)
- Deep Research (120 sources) (Mar 2026)
- UMC earnings / Silicon Analysts (Mar 2026)
- SemiAnalysis (Apr 2023)
- UMC earnings / industry reports (Mar 2026)
- TSMC Annual Report 2024 (Feb 2025)
- UMC Q1 2025 earnings / Silicon Analysts (Mar 2026)
- TSMC Q1 2025 earnings / Silicon Analysts (Mar 2026)
Related Tools
Related Analysis
Market intelligence on this data
All market intelligence →SK Hynix Commits $5.8B to Advanced HBM Packaging Facility; TSMC Confirms 2027 Wafer Price Hikes Up to 10%
SK Hynix packaging CapEx: +$5.8B committed; TSMC mature-node wafer price hike: up to +10% effective 2027; advanced-node pricing direction unquantified but directionally upward per Barchart/TSMC guidance.
China Eyes Export Controls on AI Chips & Models; AMD Helios Rack Enters Market with Azure Deal
Geopolitical: Potential restriction on TSMC fab access for Chinese-designed advanced chips (scope/timing unquantified); AMD Helios: Azure H2 2026 deployment commitment (unit volume undisclosed).
TSMC Commits $100B+ Arizona Expansion, Accelerates 3nm Conversion to Meet AI Chip Surge
Capacity: TSMC U.S. CapEx raised by +$100B incremental; total U.S. commitment reaches ~$265B per Odaily/Reuters; annual CapEx guidance lifted to $60–64B. 5nm-to-3nm fab conversion timeline accelerated (no specific % capacity figure disclosed publicly).
TSMC $265B US CapEx Commitment + ASML 30% EUV Capacity Surge Signal Multi-Year AI Supply Expansion — with Equipment Price Hike Risk
TSMC CapEx: +$100B incremental US commitment (total $265B), 2026 CapEx at high end of $52–56B range; ASML EUV/DUV capacity: +30% per year for 2 years; ASML 2026 revenue guidance raised ~15–19% vs prior midpoint (€38B → €44B midpoint); ASML EUV price hike quantum not yet disclosed — wafer/packaging cost pass-through risk flagged but unquantified.
Frequently Asked Questions
- What is current semiconductor fab utilization?
- TSMC fab utilization recovered to approximately 90% by Q1 2025, up from the cyclical trough of roughly 70% in Q2 2023. Advanced nodes (3nm, 5nm) run at near-100% utilization while mature nodes (28nm+) remain around 60–70%.
- Why did fab utilization drop in 2023?
- The 2023 utilization downturn was caused by post-COVID inventory correction in consumer electronics (PCs, smartphones), coinciding with the DRAM/NAND price crash. AI demand partially offset this at leading-edge nodes, but mature nodes saw severe underutilization.