Tech
Apple wants a second source. Intel and Samsung want it to be them.
Bloomberg reports Apple is in early-stage discussions with Intel and Samsung about producing some of its M-series chips. The talks are exploratory; the signal is significant.
Apple’s silicon strategy has, for nearly a decade, run on a single foundry relationship. Bloomberg reported on Tuesday that the company is now exploring early-stage discussions with Intel and Samsung Electronics about manufacturing some of its M-series processors, in a quiet move to diversify production away from Taiwan Semiconductor Manufacturing Company. 9to5Mac confirmed the reporting on the same day, framing it as the most concrete signal yet that Apple is taking foundry concentration risk seriously enough to act on it.
Apple is not, on the available reporting, planning to walk away from TSMC. The discussions are at an early stage, no orders have been placed, and Apple has internal concerns about whether non-TSMC technology can match the yield, performance, and timing the company has come to depend on.
The most likely scenario, by AppleInsider’s analysis, is that Apple uses Intel or Samsung for its lower-end M-series parts, the chips that ship in MacBook Air, iPad Pro, and similar mid-volume products, while keeping its highest-performance silicon on TSMC nodes. Initial shipments of any non-TSMC part are not expected until the second or third quarter of 2027.
The strategic logic is two-track. The first track is geopolitical: TSMC’s continued concentration in Taiwan is a known supply-chain risk in any scenario involving a Chinese move on the island, and Apple has been quietly diversifying around that question for years. The second track is commercial.
Intel’s foundry services have been rebuilding under Lip-Bu Tan’s leadership, with Apple as one of the customer relationships Intel has reportedly pursued most aggressively. Samsung’s foundry, while a step behind TSMC on leading-edge nodes, has historical capability and excess capacity. Both companies want Apple business badly. Apple, by extension, has unusual leverage.
The challenge nobody has fully solved
The hard problem is yield. Industry analysts at Semiwiki have tracked the gap between TSMC’s leading-edge nodes and Intel’s and Samsung’s equivalents through 2026, and the consensus is that both alternative foundries are closer to TSMC’s quality than they have been in years, but neither has fully closed the gap. For Apple, which has historically shipped tens of millions of M-series units per year and demands consistent performance across that volume, even a small yield difference compounds into meaningful product-cost and customer-experience differences.
The supply context also matters. TNW reported last week that Apple raised the entry-level Mac mini’s starting price from $599 to $799 after AI-driven demand depleted inventory at higher configurations. That kind of demand pressure makes any foundry diversification more urgent, but also riskier: a yield problem at a new manufacturing partner would amplify, not relieve, the supply constraint Apple is currently dealing with.
Tuesday’s reporting is, in itself, unlikely to change Apple’s near-term product roadmap. The longer-term signal is more consequential. Apple has, for the first time in its modern silicon era, publicly hinted that the TSMC relationship is no longer treated as singular. Whether the eventual outcome is a partial second-source arrangement, a long-running negotiating posture, or no change at all, the fact that Apple is publicly signalling diversification interest changes the bargaining table for all three foundries.
The next signal will come not from Apple but from Intel and Samsung. If either announces a leading-edge process win that includes Apple as a customer over the next 12 to 24 months, the diversification thesis will have moved from exploration to commercial reality. If neither does, the discussions will be remembered as one of the smaller chess moves in Apple’s longer game with its most important supplier.
You must be logged in to post a comment Login