TSMC's first customer for next-gen A16 process is NVIDIA, Apple skips A16 directly to A14 node

NVIDIA is TSMC's first customer for its next-gen A16 process node (1.6nm), with volume production in 2027 in Taiwan, Apple skips A16 directly to A14 node.

TSMC's first customer for next-gen A16 process is NVIDIA, Apple skips A16 directly to A14 node
Comment IconFacebook IconX IconReddit Icon
Gaming Editor
Published
2 minutes & 45 seconds read time
TL;DR: TSMC will begin production of its next-generation A16 (1.6nm) process in 2027, with NVIDIA as the first customer, while Apple plans to skip A16 for the A14 (1.4nm) node. TSMC is expanding capacity with new 2nm fabs in Taiwan and Arizona, driven by strong demand for advanced AI chips.

NVIDIA will be the first customer for TSMC on its next-generation A16 process (1.6nm) with production taking off in 2027 at its Taiwan fab, while Apple will reportedly skip A16 and use the A4 (1.4nm) process.

TSMC's first customer for next-gen A16 process is NVIDIA, Apple skips A16 directly to A14 node 80
3

We've been hearing rumors that NVIDIA would be the first customer for TSMC on its new A16 process node in 2026, but it looks like that will now take place in 2027, while other rumors said Apple was "not yet in talks" with TSMC to use its A16, with this new report suggesting the company is going directly to A14.

The new DigiTimes report also confirms that TSMC's plans to build 3 more 2nm semiconductor fabs in Taiwan in order to keep up with the ever-growing demand, with industry insiders saying TSMC's 2026 capex could reach $48-$50 billion, up from $40-42 billion this year. TSMC's advanced capacity at its Arizona fabs will also be going through some changes:

  • TSMC Arizona Plant 2: N3 mass production in 2027, and N2 later
  • TSMC Arizona Plant 3: mass production in 2029 on latest N2 and A16 nodes
  • TSMC Arizona Plant 4: construction to begin in mid-2026, will fab N2 and A16 nodes

TSMC has been incredibly busy expanding its 3nm production capacity in response to "large orders" from NVIDIA for its new 3nm AI GPUs (Blackwell, Blackwell Ultra, and next-gen Rubin chips). It was only last month that we reported that NVIDIA CEO Jensen Huang met with TSMC executives, asking for more 3nm production capacity for its chips by a large 50%.

In the middle of production capacity expansion, TSMC has constructed three new 2nm wafer fabs at the Hsinchu Science Park's Baoshan F20, Kaohsiung F22, Taichung F25, and the US plant's F21.

TSMC has been spending more money, processing its fabs faster, and expanding capacity plans are proof it has been working overtime on getting advanced nodes like 3nm and below, rapidly readjusting to customer -- NVIDIA mostly -- needs.