Apple’s next-generation A20 chip is shaping up to be noticeably more expensive than the current A19, but not nearly as costly as some earlier rumors suggested. New Taiwan supply-chain checks from Morgan Stanley point to a more balanced picture: memory costs are climbing fast, while Apple appears to be keeping its foundry costs under tighter control thanks to unusually favorable terms with TSMC.
One of the biggest wins for Apple right now is NAND flash supply. The report says Apple has locked in enough NAND inventory through the first quarter of 2026, largely due to an agreement with KIOXIA. That helps Apple protect iPhone production volumes and avoid the worst of potential shortages. However, this advantage may not last indefinitely. With the existing arrangement expected to expire, KIOXIA is likely to push for higher pricing when a new long-term deal is negotiated in early 2026.
The tougher battle is DRAM. Morgan Stanley indicates Apple is still negotiating DRAM supply, and pricing for Q1 2026 is a key point of friction. Because Apple previously benefited from long-term DRAM contracts that were especially attractive, memory makers are now expected to try to narrow the gap between those older contract rates and current spot-market pricing. The bank’s view is that Apple may only secure DRAM coverage through Q1 2026 by accepting a sequential price increase of more than 50 percent. If that plays out, DRAM becomes a major driver of cost inflation across upcoming iPhone designs and the silicon that supports them.
On the other hand, Apple appears to be getting comparatively light price increases from TSMC for leading-edge wafer production. Morgan Stanley says TSMC is expected to raise wafer prices for Apple by only low-single digits, a smaller increase than what other leading-edge customers are seeing (mid-single digits). That difference matters, especially as Apple transitions to more advanced manufacturing for its next chips.
Putting these moving parts together, Morgan Stanley estimates Apple’s A20 chips—expected to use TSMC’s 2nm process—will cost roughly 30 percent more than the A19 chips built on 3nm technology and expected to power the iPhone 17 lineup. In other words, despite the jump to a new manufacturing node, Apple’s overall chip cost increase may be kept to a level that’s significant but manageable, largely because wafer pricing pressure looks contained compared to memory pricing pressure.
This outlook also pushes back against a prior claim that the A20 could cost Apple around $280 per unit—an implied year-over-year increase of about 80 percent versus the A19. That earlier view tied the cost surge to rising memory pricing and more complex manufacturing techniques at TSMC, including first-generation nanosheet transistor technology and advanced capacitor structures in the N2P family of processes.
Morgan Stanley’s take suggests Apple’s scale still gives it powerful leverage. Even as memory suppliers attempt to reset pricing higher, Apple’s volume and strategic importance to TSMC may be helping it negotiate better wafer terms than the broader market. The end result could be a next-gen iPhone chip that’s clearly more expensive to build, but not exploding in cost the way worst-case projections implied.
If these trends hold into 2026, the key question for consumers won’t just be the performance gains from 2nm—it’ll be whether Apple chooses to absorb rising memory costs, offset them elsewhere in the bill of materials, or let higher component expenses influence iPhone pricing and configurations.
