Apple and Nvidia are considering shifting a portion of their chip production to Intel by 2028. According to Digitimes, both companies will assign non-core, low-volume orders to their compatriot, while continuing to rely on TSMC for their flagship components.
Nvidia is said to use Intel’s advanced packaging technology and is evaluating its upcoming 14A node for parts of its 2028 Feynman GPU platform, with the core GPU dies remaining at TSMC. Apple is reportedly working with Intel to produce entry-level M-series processors, reviving a hardware relationship that ended in 2020 when it transitioned to its in-house silicon.
The move reflects growing adoption of dual-foundry strategies among US chipmakers seeking to diversify production, reduce risk, comply with new domestic manufacturing mandates or simply secure capacity when there’s not a lot to go around. Companies including Google, AMD, Microsoft and Qualcomm are reportedly also evaluating their capacity allocations.
Samsung stands to benefit as well. The Korean firm’s foundry division, which is finalizing a fab in Texas, recently confirmed that it has initiated talks with major US customers. It expects 2nm orders to rise by more than 30 percent year-on-year in 2026. Last year, Samsung secured a 16.5-billion-dollar contract with Tesla, lending credibility to the notion that it’s successfully turning around a troublesome period.
Intel isn’t quite there yet. CEO Lip-Bu Tan has signaled that commitments from external customers for the 14A node will arrive in H2 2026 at the earliest. It could easily take until 2029 before these orders are turned into volume production. Apple’s interest reportedly goes out to the 18A-P sub-node, slated to debut in 2027.
For TSMC, meanwhile, the rising competition isn’t necessarily a bad thing. Offloading lower-tier production eases monopoly concerns and appeases US political demands, while reinforcing its grip on high-margin, cutting-edge chip manufacturing.
