KEY POINTS
- Intel announced risk production of its most advanced 18A-P chip node at the VLSI Symposium in Honolulu on June 16, a critical milestone toward qualifying Apple as a foundry customer.
- The deal is driven by TSMC capacity constraints as the AI boom consumes advanced packaging slots, forcing Apple to diversify its supply chain for the first time since ditching Intel CPUs in 2020.
- Traders should watch Intel's 18A-P yield data over the next two quarters and any formal announcement of production volumes tied to Apple's M7 and A21 chip designs expected by late 2027.
Intel took a decisive step toward reinventing itself as a contract chipmaker this week, announcing that its most advanced process node has entered risk production. The 18A-P node, disclosed at the VLSI Symposium in Honolulu on Tuesday, is now producing early silicon with data indicating it will meet customer specifications upon final qualification. For a company that spent the past three years and tens of billions of dollars attempting to catch Taiwan Semiconductor Manufacturing Co., this is the first tangible proof that Intel's foundry pivot may actually work.
The Apple Variable
The timing matters because of what sits on the other side of this production milestone: Apple. The Wall Street Journal reported in May that the two companies reached a preliminary agreement for Intel to manufacture certain Apple-designed processors using the 18A process. The scope covers entry-level A-series and M-series silicon, with the M7 for MacBook Pro and MacBook Air and the A21 for iPhone targeted for late 2027 production.
Intel shares surged 240% in the weeks following the initial report, and INTC remains one of the best-performing large-cap semiconductor names of 2026. But the stock's next leg depends entirely on whether 18A-P can produce chips at the yields and volumes Apple demands. Risk production is an early stage. Intel must progress through yield ramp and high-volume manufacturing qualification before any revenue materializes, and Apple is famous for walking away from suppliers that miss specs.
The strategic logic for Apple is straightforward. TSMC's advanced nodes are increasingly consumed by Nvidia, AMD, and the wave of hyperscaler custom silicon programs. Total hyperscaler capital expenditure commitments for 2026 reached $750 billion, much of it flowing to AI accelerator production at TSMC's Arizona and Taiwan fabs. That leaves Apple competing for capacity against the highest-margin customers in TSMC's history. Diversifying to Intel, even for lower-end chips, gives Apple leverage in pricing negotiations with TSMC and a hedge against geopolitical supply disruption.
What Risk Production Actually Means
Risk production sits between research and development and full manufacturing. Intel is fabricating chips on the 18A-P node and sharing results with potential customers to validate performance, power consumption, and defect density. The "P" designation refers to a performance-optimized variant of the base 18A node, which Intel has said will use its first backside power delivery architecture and Gate-All-Around transistors.
If 18A-P succeeds, Intel would leapfrog to competitive parity with TSMC's N2 node expected in volume production later this year. That is a significant if. Intel's track record on process execution over the past decade has been marked by repeated delays, most notably the years-long stumble on what was originally called 10nm. CEO Pat Gelsinger launched the IDM 2.0 strategy in 2021 with a promise to regain process leadership by 2025. Intel is now roughly 18 months behind that target.
The Foundry Revenue Question
Wall Street estimates suggest an Apple foundry contract could generate $4 billion to $6 billion in annual revenue for Intel by 2028, assuming the deal advances beyond preliminary terms. That would make Apple Intel's largest external foundry customer by a wide margin and transform Intel Foundry Services from a money-losing startup into a credible business.
But traders pricing in a full Apple relationship should note the staging. Initial volumes will likely be modest: test wafers and qualification lots through 2027, with meaningful revenue not arriving until fiscal 2028. Intel's foundry segment posted an operating loss of $5.3 billion in fiscal 2025, and narrowing that gap requires not just Apple but a broader book of external customers.
What Comes Next
The next catalyst is Intel's Q2 earnings in late July, where management will provide an update on 18A-P yields and customer engagement. Any formal announcement of Apple production timelines would likely come at Intel's Innovation conference in the fall. In the meantime, the semiconductor sector is digesting Wednesday's hawkish Fed signal, with the Philadelphia Semiconductor Index down 2.1% on the session. Intel's foundry story is a 2027-2028 thesis, but the stock trades on 2026 sentiment, and sentiment just got a jolt of cold water from Chair Warsh's first FOMC meeting.
