Intel CEO Pat Gelsinger said his company is spinning up fast to build chips for auto companies that have been impacted by the semiconductor shortage, with the idea that Intel might be able to make a difference in the medium term.
“We’re hoping that some of these things can be alleviated, not requiring a three- or four-year factory build, but maybe six months of new products being certified on some of our existing processes,” Gelsinger said. “We’ve begun those engagements already with some of the key components suppliers.” The CEO noted that the work could be done in Oregon, Arizona, New Mexico, Israel, or Ireland, which is most of the places where Intel has major manufacturing facilities.
This is an interesting argument for Intel to be making, but it’s not the whole story. Intel is talking about making an immediate pivot to help the automotive industry, and the company might have the fab space to do it. Intel was working to aggressively build out capacity starting in 2018, which gave it a bit of a leg up on this crisis. The company’s 10nm Fab 42 also came online last October, boosting production of Intel’s most advanced chips.
Production and cleanroom facilities at work in Intel’s D1D/D1X plant in Hillsboro, Oregon. (Credit: Walden Kirsch/Intel Corporation)
It can take six months for an auto manufacturer to qualify a new part for usage, after it’s been manufactured and shipped out by the foundry. The total lead time on chip manufacturing, from silicon boule to finished products in customer hands, can be 18 weeks or more.
Gelsinger doesn’t say that Intel can have final silicon shipped out to customers in just 6-9 months but that the company might be able to qualify designs in that time period. This could mean product shipments are still farther out — possibly 8-12 months away. A lot depends on when Intel started having those conversations and when it began working to qualify designs. If this is a project that started in January and we’re just now hearing about it in April, a 6-9 month shipment date from now would make sense. If Intel is just kicking things off today, that’s a very fast ramp. The auto industry might be willing to skip a few steps in the typical certification process if it means getting vehicles out the door more quickly.
Building out supply chains to cope with the silicon shortage in an enduring way isn’t just about building more foundries. We also need new packaging facilities to prepare chips for sale. Reuters has a story on the complexities of the silicon supply chain and the difficulty of bringing the entire manufacturing and assembly business back to the United States. While it’s certainly possible, it’s not going to be nearly as simple as Intel, TSMC, and Samsung building a few extra factories. We need to move diffusion, packaging, and test centers back to the United States as well.
If Intel started talking to the automotive companies before announcing its plans publicly, a 6-9 month timeline for meaningful shipments could be reasonable. If not, Gelsinger is setting a very aggressive window for getting these parts moving out the door. Call it another interesting trend to watch: Just how fast can a foundry with a limited history of working with third parties pivot to manufacturing chips for a brand-new customer with a brand-new (if fairly simple) set of chips?
As for why Intel is fairly whipping itself around (by multi-billion dollar corporate standards) to address the market? The ongoing shortage presents Intel with a rare opportunity. If the company can get its foundry business moving in earnest, it can tie its own manufacturing shift to the narrative of helping the US and the world accelerate out of the global pandemic. There’s a marketing opportunity here for any company that can help us fab our way out of the ongoing semiconductor shortage, Intel very much included.
Feature image is Intel’s Fab 42, courtesy of Intel.
Now Read:
0 Commentaires