NVIDIA has reportedly taken a closer look at Intel Foundry’s 18A manufacturing process after the companies’ widely discussed $5 billion partnership, but early evaluation didn’t produce a clear “yes” for moving forward with Intel as a production source. According to a new report citing people familiar with the matter, NVIDIA sampled Intel’s 18A node to see whether it could be a fit for future chips, then paused rather than committing to the next step.
That detail is important, but it doesn’t automatically signal bad news for Intel’s foundry ambitions. In the chip industry, sampling a process design kit (PDK) and running early tests is a normal part of qualifying any advanced node. Major fabless chip designers routinely evaluate multiple foundries and multiple process options before deciding where high-volume production should happen. A pause at this stage can mean many things: the results weren’t decisive, timelines didn’t align, or the node’s strengths didn’t match the specific type of chip being evaluated.
Intel itself has repeatedly positioned 18A as a node designed first and foremost to serve its own internal product roadmap. The company has emphasized power-efficient performance, which aligns closely with products like its upcoming Panther Lake. That makes 18A a logical fit for Intel’s own processors, where power, efficiency, and platform-level optimization are key targets.
For external customers like NVIDIA, the priorities often look different. High-performance computing and data center-class workloads demand maximum performance, high frequency headroom, predictable yields, and a mature manufacturing ramp. That’s one reason Intel’s next-generation 14A process is widely viewed as the bigger moment for winning outside customers, because it’s expected to be more directly pitched toward what large external clients want in cutting-edge, high-performance silicon.
It also matters that NVIDIA is already associated with leading-edge capacity on a competing next-generation node elsewhere, meaning Intel’s 18A may not have been the primary path for NVIDIA’s most advanced products in the first place. In that context, sampling Intel 18A could be interpreted as due diligence and strategic optionality rather than an immediate plan to shift major production.
The same report also highlights the broader backdrop around Intel’s position in the US chip manufacturing push. Ongoing discussions involving the US government have reportedly helped strengthen Intel’s standing through financial support and by elevating the company’s importance as a domestic manufacturing option. It also suggests that overseas foundries are watching closely, concerned that US policy could increasingly influence where customers choose to manufacture chips—potentially nudging more production toward Intel’s US-based capabilities.
Taken together, the takeaway is nuanced: Intel Foundry is actively trying to attract major external customers, and NVIDIA’s reported 18A sampling shows Intel is at least on the evaluation list. But the most meaningful test for Intel’s momentum with outside clients may come with the company’s 14A process, along with proof that it can consistently deliver high-quality output, strong yields, and reliable volume ramp-up at the cutting edge.
