Intel is seemingly shifting its strategy of no longer outsourcing wafers to third parties to instead embracing a longer partnership with TSMC. This could be due to the unfavourable yields from Intel's own 18A process.https://www.notebookcheck.net/Intel-reportedly-sticking-with-TSMC-as-part-of-new-long-term-partnership-strategy-amid-underwhelming-18A-yields.974177.0.html
RIP Intel
Why is this story quoting a report from last year, on 18A yields? That's an eternity in fab development.
This doesn't really mean much. Michelle, and even pat, was saying that the vast majority of Panther Lake would be on 18A. Nova Lake will have a large amount of compute dies outsourced to tsmc. They've been consistent on stating that they will always be using a mix of internal and external nodes into the future.
Yet another Intel doom and gloom article. Lazy reporting, and only serves to create speculation based on out of date information and the author's agenda. Sure, yields may not be as spectacular as they would be after a multi-year volume run, however, there are a number of other reasons Intel may outsource mobile and consumer chips. Chief among them is volume production. I've read that they are focusing their volume to Enterprise server chips and one reason to do that is they are much more profitable than consumer processors. If they can only produce enough volume for 1 million processors from their wafers then why produce something with a much lower profit margin? Intel has been pursuing an ambitious path and I hope they can succeed. I primarily run AMD chips over the years, but we need healthy competition in this space.
That article he linked was bad. Yield % means nothing, defect density is the key metric. If I remember correctly that article was accounting for some massive Broadcom chip. Partnership with TSMC for generational leaping was already happening. Crazy speculation bad reporting.