Intel up 11% on a single headline, and the headline is a big one: Google booking 3 million TPUs at Intel's foundry for 2028 production, with Nvidia reportedly evaluating the 18A node too.

Here is where I land. One marquee customer does not fix a foundry that has trailed TSMC for years. The economics of leading-edge fabs only work at scale, and a 2028 order is still a long runway with execution risk all over it. But it is the first time in a while that a credible hyperscaler has put real volume behind Intel Foundry, and that is exactly the kind of validation the bull case needed.

I would not chase the 11% candle. I want to see a second and third customer, and margin proof, before calling it a turnaround. Right now this is a story stock again, which is better than a falling knife, but it is not yet a numbers story.