Nvidia and Apple are now competing directly for TSMC wafers, as the foundry's advanced-node capacity struggles to keep pace with demand for AI accelerators, BigGo Finance reports.

The scramble marks a shift in how the industry's tightest supply constraint is priced and allocated. For years, TSMC's leading-edge lines served a handful of anchor customers with room to spare. The AI boom has erased that cushion. Nvidia's data-center GPUs consume vast slabs of leading-edge capacity, and Apple — historically TSMC's largest and most privileged customer on advanced process nodes — now finds itself sharing the queue.

Why does the bottleneck matter now?

AI chips are large, and large chips eat wafers. A single high-end accelerator can occupy a big share of a reticle, meaning each GPU shipped removes far more capacity than a phone application processor. When one customer's orders scale with hyperscaler spending, every other customer at the same node feels it. That's the dynamic now playing out between Nvidia and Apple on TSMC's most advanced process families.

The report frames the situation as a growing bottleneck rather than a crisis. TSMC continues to expand. But capacity additions land in quarters, while AI demand re-rates in weeks. The gap between those two clocks is where the friction lives.

Who gets wafers first?

Allocation at a leading-edge foundry is part contract, part leverage. Customers that co-invest in capacity, prepay, or guarantee volumes sit higher in the queue. Apple has long played that role, anchoring each new node as it ramps. Nvidia's AI business, backed by insatiable data-center demand, has given it the purchasing weight to compete for the same slices.

The result, per the report: a scramble. Both companies are pushing to lock in wafer supply, and TSMC holds the deciding hand on who gets how much.

What are the ripple effects?

• Product timelines. If wafers tilt toward AI GPUs, consumer-device roadmaps can slip or stretch.
• Pricing power. Tight leading-edge supply strengthens TSMC's hand on foundry pricing.
• Competitive dynamics. Customers without anchor status face longer waits or migration to alternative nodes and foundries.
• Supply chain concentration. More of the industry's most advanced logic sits with a single supplier, which keeps regulators and governments watching closely.

The geopolitical backdrop sharpens the commercial one. Export controls and industrial policy have pushed AI chipmakers to secure supply chains they once treated as commodities. Capacity at the leading edge has become a strategic asset, not just a procurement line item.

What comes next?

TSMC is expanding, but the report's framing suggests the squeeze tightens before it eases. As long as hyperscaler AI spending keeps scaling, Nvidia and Apple will keep bidding for the same wafer supply — and every allocation decision TSMC makes will ripple through product availability, pricing, and launch timing across the industry.