Apple's next entry-level Mac chip is quietly rewriting a three-decade taboo: the base M7 is expected to be the first Apple Silicon manufactured by Intel, built on the company's newest 18A-P process node. The deal, reported by TweakTown and TechPowerUp in May and backed by Intel's June production milestone, marks the first time Apple has handed a flagship Mac chip to its former x86 rival since the Intel-to-Apple-Silicon transition began in 2020.
Intel took a major step toward making that happen on June 16, when it announced at the VLSI Symposium in Honolulu that 18A-P has entered "risk production," the early production stage where data indicates the process will meet customer requirements at final qualification.
The M7 Roadmap Behind the Intel Deal
Bloomberg's Mark Gurman laid out the chip plan that makes Intel's win possible. Apple has compressed its silicon cycle dramatically, taping out the M7 just six months after the M6, and is skipping the M6 Pro, Max, and Ultra entirely to rush AI-focused hardware to market.
Under that accelerated schedule, the base M7 arrives in the first half of 2027 with unified memory bandwidth of about 240 GB/s, a 56 percent jump over the M5's 153 GB/s. The M7 Pro and M7 Max follow at the end of 2027, and the M7 Ultra lands in 2028 with up to 1.5 TB of unified memory and AI performance Apple says approaches Nvidia's Blackwell class.
According to the May reports, it is the base M7 MacBook chip that Intel Foundry will build on 18A-P, with future iPhone chips potentially moving to Intel's 14A node later. Apple is also reportedly weighing Intel and Samsung foundries for its A21 chips as TSMC supply constraints tighten.
Why 18A-P Matters to Intel's Turnaround
For Intel, Apple is the ultimate foundry customer. The company has struggled for years to convert its manufacturing into a competitive business for outside clients, and analysts have repeatedly said it needs a marquee win to prove its process leadership.
18A-P is the pitch. Compared with 18A, which Intel has been making at volume in Arizona since December, the node offers 9 percent higher performance or 18 percent lower power, at least 20 percent better heat resistance, and full compatibility with existing 18A buildouts.
The market is already pricing in the deal: Intel shares have more than tripled this year, after soaring 84 percent in 2025, fueled by a 10 percent U.S. government stake taken in August and Nvidia's $5 billion investment in September. CEO Lip-Bu Tan told CNBC in May he expects commitments from multiple foundry customers in the second half of 2026.
Investors reacted instantly to the Apple reports, with Intel shares popping nearly 14 percent in May on news of a preliminary deal. Chip analyst Ben Bajarin told CNBC that Apple is likely to wait for 18A-P before committing production.
What Intel must now prove:
- Yield: Counterpoint Research's Neil Shah says committing to more than 90 percent yield in the first month could attract more customers.
- Arm expertise: Intel has built its volume on x86, while Apple, Google, and Amazon chips run on Arm, an architecture Intel has not mass-produced.
- Packaging: Intel's EMIB technology rivals TSMC's CoWoS, and analysts call packaging the lowest-hanging opportunity while TSMC faces bottlenecks.
TSMC is not standing still. The market leader is expanding a $165 billion campus just 50 miles north of Intel's Arizona plant, and it is expected to keep the bulk of Apple's processor volume even if the M7 deal closes.
For Apple, the motivation is diversification and supply security. The memory chip shortage has already forced Apple to pull the 128 GB Mac Studio this year, and the M7 Ultra's 1.5 TB configuration depends on DRAM availability easing by 2028.
The biggest questions now are timing and yield. If 18A-P qualifies on schedule, the base M7 in the first half of 2027 becomes a landmark: Apple Silicon built by Intel, at a moment when both companies are betting their AI futures on it.
Comments