Semiconductor Showdown: Apple and NVIDIA Battle for TSMC’s A16 Capacity

As AI demand surges, Apple and NVIDIA are locked in a fierce bid for TSMC’s forthcoming angstrom-class A16 capacity, alongside the A14 process, according to information obtained by Digitimes. The piece notes that TSMC is expanding its advanced-node capacity to meet the needs of major clients such as Apple and NVIDIA, with plans to accelerate 3nm ramp-ups at its STSP fabs and to pursue further angstrom-class nodes.

Renowned analyst Ming-Chi Kuo is cited as saying Apple intends to mitigate supply-chain risk and may explore a partnership with Intel on its 18AP process, potentially enabling Intel Foundry to ship Apple chips as early as 2027 for entry-level M-series processors. However, the article makes clear that Apple would still reserve TSMC’s most valuable advanced-node capacity for high-performance product lines.

The piece also observes that long-time partners such as AMD and MediaTek are expected to vie for these slots, and that Intel’s involvement would be limited in scope for now. For Intel, landing Apple would boost its foundry credibility, while for Apple, a diversified supply chain could strengthen its bargaining position with TSMC.