Intel's 18A-P process delivers 9% faster chips or 18% lower power, giving its foundry a technical argument to win external customers.
Intel's 18A-P process delivers 9% faster chips or 18% lower power, giving its foundry a technical argument to win external customers.

Intel's 18A-P process delivers 9% faster chips or 18% lower power, giving its foundry a technical argument to win external customers.
Intel began risk production of its 18A-P manufacturing process, offering chips that run 9% faster at the same power or cut energy use by 18%, as the company seeks external customers to offset billions in quarterly foundry losses.
"Our updates at VLSI signal to Intel Foundry customers and partners that we are fully committed to leading edge process innovation over the long term," Naga Chandrasekaran, executive vice president and general manager of Intel Foundry, said at the 2026 VLSI Symposium.
The 18A-P variant introduces Power Boost, a dual-contact transistor option that increases drive current and frequency at matched capacitance. Thermal resistance improved 20% to 40% through materials and design changes, while via resistance dropped 10% to 30% using geometric and materials optimizations. The process is fully design-rule compatible with standard Intel 18A, allowing straightforward reuse of existing intellectual property and design flows.
The milestone comes as Intel's foundry business, which lost billions last quarter, tries to convert a $5 billion investment from Nvidia into a manufacturing commitment and advance exploratory discussions with Apple. Counterpoint Research analysts said Apple will likely test 18A-P for its next-generation M7 processor, but matching Taiwan Semiconductor Manufacturing Co.'s manufacturing yield will be the deciding factor.
Intel presented the 18A-P details alongside longer-range research including monolithic CFET (complementary FET) inverters at a 45nm gate pitch, 300mm monolithic integration of gallium nitride power devices with silicon logic, and subtractive ruthenium interconnects with airgap integration that achieved roughly 35% capacitance reduction versus copper.
The company also shared silicon results from CPU cores built on its gate-all-around transistor and backside power delivery architecture — technologies it brought to market last year with the standard Intel 18A node. Those cores demonstrated roughly 30% frequency improvement at low voltage (about 0.5V), alongside an 11% routed area reduction and 10X dynamic voltage droop reduction, according to Intel Fellow Eric Karl.
Who Wins, Who Loses
For Intel, 18A-P represents a technical proof point that its process roadmap is advancing on schedule after years of delays that cost it market share and credibility. The company is targeting major external customers to fill capacity at its fabs and spread the enormous fixed costs of leading-edge manufacturing.
For TSMC, Intel's progress introduces a credible alternative for customers seeking geographic diversity in chip production. TSMC currently manufactures chips for Apple, Nvidia, and most of the semiconductor industry on its 3nm and upcoming 2nm nodes. Intel's 18A-P competes directly with TSMC's N2 process, though Intel has not disclosed independent benchmark comparisons.
Apple stands as the most consequential potential customer. A win with the M7 processor would validate Intel's foundry strategy and provide a reference account that could attract additional design wins. But Counterpoint Research cautioned that yield — the percentage of functional chips from each wafer — matters more than node specifications. TSMC's manufacturing consistency, honed over decades serving Apple's exacting quality standards, sets a high bar.
Investment Angle
Intel shares rose 3% to $120.60 in premarket trading Wednesday, recouping some of the prior session's 8.5% decline during a broader technology selloff. The stock trades at roughly 22x forward earnings, a discount to Nvidia at 35x and TSMC at 28x, reflecting the market's skepticism about Intel's foundry turnaround. A confirmed Apple design win on 18A-P could trigger a re-rating; failure to match TSMC's yields would further damage credibility and likely push the stock lower.
This article is for informational purposes only and does not constitute investment advice.