Recent data from TSMC’s supplier reports indicate that Nvidia has surpassed Apple as the foundry’s largest customer, driven by record demand for AI GPUs. According to statements by Nvidia CEO Jensen Huang, the AI chipmaker’s monthly wafer starts have climbed by more than 40% year-over-year, eclipsing Apple’s long-standing leadership. This shift marks the first time since 2016 that Apple has not held the top spot for wafer volume at TSMC.

For years, Apple benefited from an exclusive arrangement with TSMC that allowed priority access to new process nodes and waivers on defective die charges. Internal supply chain documents reviewed by analysts show that Apple’s allocation at the 3 nm and 4 nm nodes is being reduced by up to 25% compared with 2025, while TSMC’s standard defect chargebacks will now apply. Industry consultants warn that this loss of perks could add between 5% and 8% to Apple’s per-chip fabrication costs on advanced nodes.

In response to capacity constraints and higher wafer prices at TSMC, Apple is in advanced discussions to outsource production of its lower-end M-series processors to Intel’s contracted fab operations. Citing a note from analyst Ming-Chi Kuo, these talks envisage an annual run rate of 15 million to 20 million chips on Intel’s 18A process beginning in late 2027. Apple is also evaluating Intel’s forthcoming 14A node for next-generation iPhone–class SoCs, with volume production slated for 2028.

The potential diversification of Apple’s foundry network represents a strategic pivot to secure capacity and control manufacturing costs. Should Apple finalize orders with Intel, the company would reduce its dependence on a single supplier while potentially locking in more favorable wafer pricing. Investors should monitor updates on Intel’s process yields and TSMC’s capacity guidance, as these developments will directly influence Apple’s gross margin outlook and supply chain resilience through 2029.