Intel Starts Risk Production of 18A-P, Its Most Advanced Chip Node, Ahead of Apple Foundry Deal
- Intel's 18A-P node entered risk production and was detailed at the VLSI Symposium in Honolulu on June 16 [1]
- The process delivers 9% higher performance at iso-power, 18% lower power at iso-performance, and 50% better thermal conductivity versus base 18A [2]
- Apple and Intel reached a preliminary foundry agreement in May 2026, with initial chip shipments not expected until Q2/Q3 2027 [3]
- Intel's base 18A process has been in volume production at Fab 52 in Chandler, Arizona since December 2025 [4]
- INTC shares traded around $117 on June 16, with a market cap near $588 billion, up from a 52-week low of $18.97 [5]
Intel said on Tuesday that it has begun risk production of 18A-P, the enhanced variant of its most advanced chip manufacturing process, a step the company framed as a path toward securing major foundry customers — most notably Apple. The announcement was made at the VLSI Symposium in Honolulu [1].
The 18A-P node delivers 9% higher performance at the same power level — or 18% lower power at the same performance — compared to the base 18A process that Intel has been manufacturing at volume since December 2025. Intel also reported 50% better thermal conductivity and 30% tighter process variation in the enhanced node [2].
The milestone is significant because 18A-P is widely understood to be the target process for a preliminary manufacturing agreement between Intel and Apple, first reported in May 2026. Under that deal, Intel would produce some Apple Silicon chips at its U.S. facilities, making it a secondary foundry supplier alongside Taiwan Semiconductor Manufacturing Co. Initial shipments are not expected until the second or third quarter of 2027 [3].
What 18A-P Brings to the Table
Intel positioned 18A-P as a meaningful upgrade over the base 18A node rather than a minor revision. Beyond the headline performance and power improvements, the company added new high-performance and low-power RibbonFET transistor variants, expanded from four logic threshold-voltage pairs to five or more, and achieved 30% tighter skew corners — a measure of manufacturing consistency that directly affects yield [2].
Critically, 18A-P retains the same contacted poly pitch and library heights as base 18A, meaning chip designs built for the original process can be ported to 18A-P with relative ease. That backward compatibility lowers the barrier for foundry customers already designing on 18A to upgrade [2].
The base 18A node has been in high-volume manufacturing at Intel's Fab 52 in Chandler, Arizona since December 2025, where it powers the company's Panther Lake processor family for AI PCs. Intel described yield stabilization at Fab 52 as a key milestone in its manufacturing turnaround [4].
The Apple Connection
Apple and Intel Foundry Services reached a preliminary agreement in May 2026 for Intel to manufacture some Apple Silicon processors, according to reports from the Wall Street Journal and CNBC. Intel's stock surged 13.9% on the day the deal was first reported [3].
The arrangement would mark Apple's first use of a non-TSMC foundry for its custom processors since the company began designing its own chips in 2020. Apple has historically relied exclusively on TSMC for advanced iPhone, iPad, and Mac processors, and has made multibillion-dollar investments in TSMC's Arizona facility [3].
Specific details — including which Apple product lines would use Intel-made chips, whether Apple would provide upfront funding, and the total contract value — have not been disclosed. Analyst Ming-chi Kuo predicted in November 2025 that Intel would supply processors for iPad Pro and entry-level MacBook Air models [3].
Intel's Foundry Momentum
Intel CEO Lip-Bu Tan has been publicly building the case that Intel's foundry business is approaching an inflection point. In a May 2026 interview with CNBC, Tan said the company had 'multiple customers engaged' and expected firm commitments in the second half of the year [6].
Beyond Apple, Intel has reportedly signed foundry deals with Tesla and Google, though the scope and volume of those agreements remain unclear. Tan recounted one unnamed customer requesting a 3x increase in their production forecast, offering advance payments to secure capacity [6].
The foundry push is central to Intel's strategic transformation under Tan, who took over as CEO after Pat Gelsinger's departure. Intel's stock has recovered dramatically from its 52-week low of $18.97 to trade around $117, giving the company a market capitalization near $588 billion [5].
What's Next
Risk production is an early manufacturing stage in which Intel produces chips using the 18A-P process to validate that it meets customer requirements before moving to volume production. The transition from risk production to high-volume manufacturing typically takes several quarters [1].
For the Apple relationship, the key near-term milestone is the formal signing of a binding production contract. Current reporting describes the agreement as preliminary, with volume shipments targeted for Q2 or Q3 of 2027. Apple has already reserved substantial capacity on TSMC's upcoming 2nm node, suggesting any Intel arrangement would initially supplement rather than replace its primary foundry relationship [3].
Intel's ability to demonstrate competitive yield and reliability on 18A-P will be the decisive factor in whether Apple — and other potential foundry customers — commit to large-scale orders. The VLSI presentation served as Intel's most detailed public case yet that the process is ready for prime time [2].
Further sources
The stories that matter, in one email. Free — unsubscribe anytime.