Nvidia Isn’t Committing To 18A. What It Means For Intel Stock

Intel (NASDAQ:INTC) foundry comeback story hit a speed bump a few days ago after reports surfaced that AI titan Nvidia (NASDAQ:NVDA) was pausing testing on Intel’s most advanced manufacturing process, Intel 18A. The news sent Intel shares down as much as 4% in early trading, before they partially recovered later in the day. While Intel says its roadmap is “progressing well,” the hesitation from a potential customer like Nvidia refocuses the debate on a single question: is Intel 18A ready for leading-edge workloads?

Image by Michael Schwarzenberger from Pixabay

Stock picking can fail no matter how good the strategy is. High Quality Portfolio turns single-stock insights into a robust market beating portfolio strategy.

What Intel 18A Is, and Why It Matters

Intel 18A refers to an 18-angstrom, or roughly 1.8-nanometer, manufacturing process and is Intel’s most advanced node to date. The process introduces RibbonFET gate-all-around transistors and PowerVia backside power delivery, technologies designed to improve performance and power efficiency while enabling smaller transistor geometries.

The node is intended to place Intel slightly ahead of TSMC’s N2 process, which operates at a 2-nanometer node and entered volume production in late 2025. More importantly, 18A is the foundation of Intel Foundry Services and the company’s primary attempt to re-establish credibility at the leading edge.

Why Nvidia’s Pause Matters

Intel is separating its Product and Foundry businesses, and the foundry unit needs big, high-volume customers to justify its capital investments. Nvidia would have been an ideal anchor tenant, as its AI accelerators sit at the extreme end of density, power, and yield sensitivity. Nvidia’s rapid growth, with revenues on track to rise more than 60% this year, also increases its exposure to Taiwan-centric supply chains. That makes diversification strategically attractive, positioning Intel as a potential geopolitical hedge and a win-win partner if execution aligns.

That being said, a pause in testing may not imply rejection, but it does suggest that Intel 18A has not yet met Nvidia’s internal production thresholds. Current estimates place 18A yields in the mid-50% to mid-60% range. That level is probably sufficient for Intel’s own CPUs and for early external designs, but it may fall short for high-volume AI accelerators that demand both high yields and tight variability control.

For Nvidia, yield consistency is non-negotiable. With AI GPUs priced at tens of thousands of dollars, even small yield variability can materially affect margins and supply reliability. Internal processors and custom silicon can tolerate learning-curve inefficiencies because volume ramps are flexible. Nvidia’s flagship GPUs – which are in extremely high demand – cannot. Overall, the issue is not whether 18A works, but whether it works predictably at scale.

18A Is Already Being Used

Despite the Nvidia headline, Intel 18A is already entering high-volume manufacturing at its Fab 52 in Arizona. This remains a meaningful milestone.  Intel is the first major user, ramping production of Panther Lake laptop CPUs and Clearwater Forest data center processors, both built on 18A. These products are essential for refining yields and validating process consistency.

Externally, Intel still has committed customers. Microsoft plans to use 18A for its Maia AI accelerators, and Amazon has signed a multi-year agreement for custom silicon through Intel Foundry. These designs are probably less yield-sensitive than Nvidia’s flagship GPUs but still require leading-edge performance. Reports also suggest Apple may use Intel for select chips, potentially at the lower end of its M-series lineup, which would further support utilization and learning curves.

The Bottom Line

Nvidia’s pause is a reminder that Intel’s foundry recovery hinges on manufacturing execution, not ambition. Intel 18A is real, in production, and already supporting multiple products. The process is not failing, but it is still maturing. The key question for investors is whether Intel can convert technical capability into repeatable, high-confidence yields that persuade top-tier customers to commit volume. This is crucial to support Intel’s stock price, which has rallied by close to 80% this year.

The Trefis High Quality (HQ) Portfolio, with a collection of 30 stocks, has a track record of comfortably outperforming its benchmark that includes all three – the S&P 500, S&P mid-cap, and Russell 2000 indices. Why is that? As a group, HQ Portfolio stocks provided better returns with less risk versus the benchmark index; less of a roller-coaster ride, as evident in HQ Portfolio performance metrics.

Invest with Trefis Market-Beating Portfolios
See all Trefis Price Estimates