Intel Foundry Boosts Equipment Orders 50%+ as 14A Node Draws Apple, Nvidia, and AMD Interest

By NineScrolls Team · 2026-04-21 · 5 min read · Industry

Equipment Orders Surge 50%+ Year-Over-Year

Intel Foundry has ramped chipmaking equipment orders by more than 50% year-over-year since the start of 2026, according to a TrendForce report published April 20. The surge signals that CEO Lip-Bu Tan is backing his foundry turnaround strategy with real capital, reversing months of uncertainty about whether Intel would follow through on its most advanced process nodes.

Taiwanese equipment suppliers are among the early beneficiaries. E&R Engineering, which supplies Raman inspection, laser processing, and plasma-related tools, has seen increased order volumes from Intel. KINIK Company, a provider of diamond conditioning discs for chemical mechanical planarization (CMP), has been supplying Intel since the second half of 2025. The order acceleration comes as Intel prepares infrastructure for both its 18A ramp and the buildout of 14A capacity.

14A Node: High-NA EUV, RibbonFET, and PowerDirect

Intel's 14A process — its 1.4-nanometer-class node — will be the company's first to deploy High-NA EUV lithography from ASML, alongside second-generation RibbonFET gate-all-around (GAA) transistors and PowerDirect, an upgraded version of the backside power delivery technology introduced with 18A. The node also introduces Turbo Cells, an optimization technique for improving timing paths without significant area or power trade-offs.

CEO Tan declared in a recent briefing: "We are going big time into 14A. Stay tuned, we are going to see a lot of great momentum on the 14A in terms of yields and IP portfolio to serve the customer well." This represents a sharp reversal from July 2025, when Tan suggested Intel might pause or discontinue 14A development if it failed to secure external customers.

The Process Design Kit (PDK) 0.5 has already been released, and PDK 1.0 — the version that enables customers to begin serious design work — is expected this fall. Production readiness is targeted for 2027.

Customer Pipeline: Apple, Nvidia, AMD, and Google in Evaluation

Two prospective clients are actively evaluating 14A test chips, according to Tan, who expects firm supplier decisions in the "second half of this year, extending into the first half of 2027." No fully committed external 14A customers exist yet, but the evaluation list includes some of the industry's largest chip designers.

Apple is reportedly considering Intel 14A for lower-end A-series iPhone processors, with a potential production timeline around 2029, as well as certain M-series chips for MacBooks and iPads. AMD and Nvidia are evaluating the node for next-generation server CPUs and GPUs. Google and other hyperscalers are interested as they seek alternatives to TSMC's increasingly constrained capacity.

UBS raised Intel's price target to $65, citing the 14A PDK release as a near-term catalyst. However, Intel's current capital expenditure plan lacks dedicated 14A capacity funding for third-party clients — landing major external orders would require substantial additional investment before revenue materializes, further testing Intel Foundry's path to profitability.

18A Manufacturing Progress and Yield Ramp

Intel 18A entered high-volume manufacturing in October 2025, making it the first process node to combine RibbonFET gate-all-around transistors with PowerVia backside power delivery. Yields have been improving at roughly 7% per month, though CFO David Zinsner has indicated that cost thresholds will not reach desired levels until the end of 2026 at the earliest.

The node is now producing Panther Lake (Core Ultra Series 3 laptop processors) and Clearwater Forest, a Xeon server CPU with up to 288 Darkmont efficiency cores using EMIB 3.5D and Foveros Direct 3D packaging at 45-micron bump pitch. Intel's Fab 34 in Ireland handles volume production on Intel 4, while the D1 facility in Hillsboro, Oregon, supports development.

The Terafab project — a $25 billion joint venture between Tesla, SpaceX, and xAI with Intel as primary foundry partner — is slated for test production by 2029. Applied Materials, Tokyo Electron, and Lam Research have all received requests for price quotes and delivery timelines on etchers, depositors, cleaning devices, and testers for the megaproject.

What This Means for Plasma Processing and Thin Film Deposition

Intel's 50%+ equipment order surge is a direct demand signal for plasma processing and thin film deposition equipment makers. The 14A node's gate-all-around architecture requires significantly more deposition and etch steps than planar or FinFET transistors. RibbonFET channels demand atomic-layer-precision films — typically deposited by ALD or plasma-enhanced ALD — with tight uniformity across the nanosheet stack. Each additional nanosheet layer in a GAA device adds multiple plasma etch steps to release the channels and pattern the gate wrap-around.

PowerDirect backside power delivery introduces an entirely new set of deposition and etch requirements on the wafer's rear side. Through-silicon via (TSV) formation for backside contacts relies on deep reactive ion etching (DRIE) and conformal barrier/seed layer deposition by PVD or CVD. The front-side routing, now freed from power rails, shifts to finer-pitch copper or molybdenum interconnects that require advanced PECVD dielectric films and precise plasma etch for dual-damascene patterning.

For the equipment supply chain, the Intel ramp translates to increased demand for plasma sources, RF generators, mass flow controllers, vacuum pumps, sputtering targets (particularly molybdenum and ruthenium for advanced metallization), and high-purity gas delivery systems. The integration of High-NA EUV at 14A also drives demand for thinner, more precisely deposited photoresist underlayers and hard masks — materials typically applied by PECVD or spin-on processes with subsequent plasma treatment. Equipment suppliers serving the etch, deposition, and process monitoring segments stand to benefit as Intel scales from development to volume production across two leading-edge nodes simultaneously.