TSMC's $49 billion 1.4nm wafer fab is running ahead of its original construction timeline, lifting US-listed shares more than 3 percent in pre-market trading.
TSMC's $49 billion 1.4nm wafer fab is running ahead of its original construction timeline, lifting US-listed shares more than 3 percent in pre-market trading.

TSMC's $49 billion 1.4nm wafer fab is running ahead of its original construction timeline, lifting US-listed shares more than 3 percent in pre-market trading.
TSMC's $49 billion 1.4nm wafer fab is ahead of its original construction schedule, lifting US-listed shares more than 3 percent in pre-market trading as the foundry accelerates its most advanced node yet.
The accelerated timeline, reported by Taiwan's Economic Daily News, follows TSMC's disclosure that its 2nm process has logged four times as many tapeouts as the 3nm node at the same stage, with the company targeting 100,000 monthly 2nm wafers by the end of 2026.
The 1.4nm fab sits alongside a broader expansion: Fab 20's 2nm line has already reached 20,000 monthly wafers, and the 3nm process is running at 180,000 monthly wafers — ahead of its Q4 2026 target — driven by demand from NVIDIA, AMD, and Broadcom. The 2nm node contributed 3 percent of TSMC's Q3 2026 revenue.
The early completion matters because Apple, NVIDIA, AMD, and Qualcomm are all queuing for advanced node capacity. Apple has signaled it may move to 1.4nm after just two generations of 2nm to secure supply, with TSMC targeting mid-2028 for 1.4nm production — a timeline that now appears achievable ahead of schedule.
The 1.4nm acceleration builds on a 2nm ramp already straining capacity. Fab 20, TSMC's first dedicated 2nm facility, went from 20,000 monthly wafers to a projected 100,000 within four months — a fivefold expansion driven by AI demand. Apple's A20 Pro, expected in the iPhone 18 Pro and Pro Max this September, will be among the first consumer chips on the node, followed by Qualcomm and MediaTek SoCs powering 2027 flagships.
On the data center side, AMD's MI455X GPUs and NVIDIA's Rubin Ultra are both slated for 2nm production, competing for the same wafer supply. TSMC's 3nm line has not seen demand wane either — the company reached 180,000 monthly wafers ahead of schedule, with NVIDIA, AMD, and Broadcom absorbing the output. The 2nm node already contributed 3 percent of TSMC's Q3 2026 revenue, a figure that will climb sharply as more customers tape out designs.
The demand pressure explains why TSMC is moving so aggressively on construction. The 2nm node (which packs more transistors per square millimeter, improving performance per watt) is expected to be the primary revenue driver through 2027, and the 1.4nm node will follow as the next major transition. Apple's reported plan to move to 1.4nm after just two generations of 2nm — rather than the typical three to four — reflects the supply constraints that AI demand has created across the industry.
The construction acceleration puts additional pressure on Samsung Foundry and Intel, both racing to close the gap in advanced nodes. Samsung's 2nm GAA process has faced yield challenges, while Intel's 18A node has yet to secure a major external customer. TSMC's ability to bring 1.4nm online early would extend its manufacturing lead into the late 2020s, potentially locking in Apple, NVIDIA, and AMD as long-term customers.
TSMC shares, which have rallied on the back of AI infrastructure spending, now face a supply-constrained environment where the company's biggest challenge is building capacity fast enough to meet demand. The 1.4nm fab's early completion is a direct response to that pressure. With the 2nm node expected to be the primary revenue driver through 2027 and 1.4nm following in 2028, TSMC's execution on construction timelines will determine whether it can sustain its roughly 60 percent share of the global foundry market.
This article is for informational purposes only and does not constitute investment advice.