What are the key details and timeline of TSMC's accelerated 3nm and 2nm production ramp in 2026, including wafer start targets, customer demand from AI giants like NVIDIA, AMD, and Broadcom, expansion plans with new fabs in Taiwan, Arizona, and Japan, and the outlook for next gen
TSMC is executing its most aggressive capacity expansion in history in 2026, driven by insatiable AI demand from NVIDIA, AMD, Broadcom, and Apple. Both 3nm and 2nm are fully booked through end of 2026, with wafer start targets raised well above original plans [1][5].
What are the key details and timeline of TSMC's accelerated 3nm and 2nm production ramp in 2026, including wafer start targets, customer demAI-generated editorial hero image for What are the key details and timeline of TSMC's accelerated 3nm and 2nm production ramp in 2026, including wafer start targets, customer dem.
AI Prompt
Create a landscape editorial hero image for this Studio Global article: What are the key details and timeline of TSMC's accelerated 3nm and 2nm production ramp in 2026, including wafer start targets, customer dem. Article summary: TSMC is executing its most aggressive capacity expansion in history in 2026, driven by insatiable AI demand from NVIDIA, AMD, Broadcom, and Apple.. Topic tags: general web, ai, workflow, api, growth. Style: premium digital editorial illustration, source-backed research mood, clean composition, high detail, modern web publication hero. Use reference image context only for broad subject, composition, and topical grounding; do not copy the exact image. Avoid: logos, brand marks, copyrighted characters, real person likenesses, fake screenshots, UI text, readable text, watermarks, charts with fake numbers, clickbait thumbnails, icons, and tiny thumbnail layouts. Mak
openai.com
TSMC is executing its most aggressive capacity expansion in history in 2026, driven by insatiable AI demand from NVIDIA, AMD, Broadcom, and Apple. Both 3nm and 2nm are fully booked through end of 2026, with wafer-start targets raised well above original plans .
3nm Production Ramp
New target: 180,000 wafer starts per month (WPM) — up 20% from the original plan of 150,000 WPM .
Timeline accelerated: TSMC is expected to hit 180K WPM by early Q4 2026, roughly 2–3 months ahead of earlier expectations .
YoY growth: This represents over 40% year-on-year capacity increase from 2025 .
Studio Global AI
Search, cite, and publish your own answer
Use this topic as a starting point for a fresh source-backed answer, then compare citations before you share it.
What is the short answer to "What are the key details and timeline of TSMC's accelerated 3nm and 2nm production ramp in 2026, including wafer start targets, customer demand from AI giants like NVIDIA, AMD, and Broadcom, expansion plans with new fabs in Taiwan, Arizona, and Japan, and the outlook for next gen"?
TSMC is executing its most aggressive capacity expansion in history in 2026, driven by insatiable AI demand from NVIDIA, AMD, Broadcom, and Apple.
What are the key points to validate first?
TSMC is executing its most aggressive capacity expansion in history in 2026, driven by insatiable AI demand from NVIDIA, AMD, Broadcom, and Apple. Both 3nm and 2nm are fully booked through end of 2026, with wafer start targets raised well above original plans [1][5].
What should I do next in practice?
3nm Production Ramp New target: 180,000 wafer starts per month (WPM) — up 20% from the original plan of 150,000 WPM [1][5].
Key customers: NVIDIA, AMD, and Broadcom are aggressively securing orders, alongside Apple . Omdia confirms TSMC's 2nm and 3nm nodes are "largely booked" by these AI leaders .
Wafer pricing: 3nm N3/N3E wafers are priced at approximately $19,500 per wafer as of early 2026, with lead times exceeding 50 weeks .
2nm Production Ramp
Mass production start: 2nm (N2) entered mass production in Q4 2025, transitioning from FinFET to nanosheet (GAA) transistor architecture .
Year-end 2026 target: TSMC aims to reach 100,000 WPM for 2nm by end of 2026 .
CAGR guidance: TSMC expects 2nm capacity to grow at a compound annual rate of 70% from 2026 through 2028 .
Customer lineup: AMD is targeting 2nm-based CPU production beginning in 2026; Google and AWS are also aiming for early 2nm adoption . NVIDIA is expected to move its AI accelerator orders to 2nm in volume .
Yield: The 2nm yield learning curve is outperforming the 3nm predecessor at the same stage .
Fab Expansion Plans
Taiwan (home base):
Up to 10 fabs are reportedly under construction or starting in 2026 across Taiwan's northern, central, and southern science parks .
Five 2nm fabs are simultaneously ramping this year — two in Hsinchu (Fab 20) and three in Kaohsiung (Fab 22) — the most aggressive single-node buildout in TSMC's history .
TSMC doubled its historical construction pace, building or converting nine fab phases annually in 2025–2026, up from a historical average of four .
Arizona, USA:
TSMC announced an additional $100 billion investment in Arizona in July 2026, bringing total planned investment to $265 billion .
Fab 1 (4nm): Already ramping production and contributing to profits .
Fab 2: Equipment move-in scheduled for Q4 2026, with mass production advanced to H2 2027 (from 2028) .
Fab 3: Equipment move-in targeted for September 2027 .
TSMC is acquiring a second land parcel in Arizona to establish an independent "gigafab cluster" .
Japan (Kumamoto):
TSMC announced plans in February 2026 to manufacture 3nm chips in Kumamoto, with an investment reported at approximately $17 billion .
Next-Generation 1.4nm (A14) Outlook
Fab location: Taichung Science Park (Central Taiwan).
Construction status: Ahead of schedule. The first building (Phase 1) is expected to be completed before April 2027 .
Trial production: Could begin as early as Q3 2027 .
Mass production target: Mid-2028 (2H 2028) .
Investment: Estimated $48.5–$49 billion for the 1.4nm facility .
Technology: Will use 2nd-generation GAAFETs (NanoFlex Pro) with current EUV lithography — TSMC has told ASML it will not adopt High-NA EUV before 2029 .
Customer interest: Strong interest from Apple, Qualcomm, NVIDIA, AMD, and AI/HPC customers; internal A14 validation is already nearing 90% of target .
Longer-term roadmap: After A14 (1.4nm), TSMC expects A13 and A12 to both reach mass production in 2029 .