How CEO Lip-Bu Tan gave Intel a lifeline after meeting with Trump in August; sources say Nvidia tested Intel's 18A to make chips but stopped moving forward
Reuters
Context & Ripple Effects
Intel's position was already under strain: its board had reportedly slowed Tan's efforts to raise capital and pursue an AI acquisition, while Tan was preparing an August meeting with Trump on potential U.S.-Intel cooperation. The meeting therefore became a pivotal part of management's effort to preserve strategic options.
The reported outcome of Nvidia's 18A evaluation is consequential because a test that does not advance leaves Intel without the customer validation that a leading chip designer could have provided for its manufacturing push.
First-order effects
Intel receives a near-term lifeline tied to Tan's political engagement, but Nvidia's decision not to move ahead means 18A does not gain Nvidia as a committed manufacturing customer.
Nvidia can keep its chip-production plans separate from Intel's 18A process rather than making an immediate shift to Intel's fabs.
Second-order effects
Intel must seek other customers or technical milestones to demonstrate that 18A can win external production business; prospective customers may treat Nvidia's non-progression as an execution-risk signal.
Political support becomes more important to Intel's strategic flexibility as commercial foundry validation remains unsettled, increasing scrutiny of how public and corporate interests shape its recovery.
Third-order effects
If major chip designers continue to test alternative foundries without committing volume, second-source capacity will remain difficult to establish: technical qualification alone will not be enough to redistribute manufacturing demand.
The episode points to a more policy-linked semiconductor industry, where domestic manufacturing objectives can sustain strategic capacity even when customer adoption is not yet decisive.
The trend: This is one data point in the push to build second-source chip manufacturing capacity, with customer qualification and government backing both determining whether new foundry options become viable.
Intel has been clearly stating for months that major customers were unlikely to adopt 18A due to yield and PDK issues. The pursuit of big customers has shifted to 14A, which has shown better PDK progress. Framing Nvidia not using 18A as incremental news without providing that c…
@PatrickMoorhead 100% 💪🏻 It is well known 18A is internal as well as that most fabless explored the node and decided it wasn't ready for prime time. 👊🏻
@danielnewmanUV @PatrickMoorhead I think it's more the fact that NVIDIA was creating test chips there through a government funded program explicitly, not an actual committment to anything ever.
Re NVIDIA no longer making test chips on 18A. IT WAS A GOVERNMENT PROGRAM to allow potential partners to build designs using PDKs. It was never a committment to anything longer term than a few tests. I've been telling investors and others for ages and still, no-one seemed to r…
Hard to keep track I get it, but Intel had already said 18A was ceasing as a third party customer node and was moving efforts to 14A. Nothing burger of news.
Odd take that seems very dated to me. The company characterized months ago that 18A was retargeted to internal node. 18AP is a much more appropriate node for what $NVDA needs. And 14A, of course. $INTC
According to Reuters, NVIDIA has halted its testing of Intel's 18A manufacturing process. Sources also say that Intel CEO Lip-Bu Tan's dealmaking capabilities are serving as a catalyst for the company's turnaround. A U.S. Commerce Department official said the United States has …
“The deal gave Intel a too-strategic-to-fail aura and opened doors to potential partners who might want to win the president's favor. It also may pave the way for the government to take more equity stakes in businesses the administration deems strategic” www.reuters.com/world/us…