You are currently viewing SemiWiki as a guest which gives you limited access to the site. To view blog comments and experience other SemiWiki features you must be a registered member. Registration is fast, simple, and absolutely free so please, join our community today!
Reflecting, a destructive paranoia began around 1990 when Intel was under siege. AMD was attacking when they weren't licensed for the 486. RISC architecture appeared ascendant. As an image and a rallying point Andy gave us the "Fortress Intel" graphic where defensive and preservation actions...
Key Intel 18A deliverable is a flawless PantherLake ramp. This shows existence proof of decent yields and manufacturing capability. That will be a message that Intel is once again real. Can’t imagine anyone signing for 14A without that demonstration.
I suspect 18A will be successful, but he saddled the company with $41B in private equity debt and another $20B "build it and they will come" space expansion. That I think will keep him from hero consideration.
Could be that because of the great transparency employees learned/understood how to handle sensitive information. Intel is not transparent internally, people go into vapor-lock when info is shared, and most of the sensitive communications end up getting leaked.
Internal memos, keynotes, and meetings for upper management are all "nice", but wrt to the 100,000+ employees he and KOB are largely MIA and it is not clear where the ship is being steered. Are employees optimistic? Those I know...not so much.
Pat had a distinct inability to attract and identify talent, Naga being the exception. Lip-Bu is wasting no time clearing ELT and +1 obstacles. What I expect will prove more interesting though is the "right-sizing" of the portfolio and employee populations. The pieces spun off and programs...
Quite possibly, or maybe simply "I can do it." Pat, in my opinion, had the right idea, but with his ELT hires/appointments and actions demonstrated near zero idea on how to make it happen.
Hotard was one of several less than stellar Pat additions or appointments to the Intel ELT. No loss. Intel will be well-served to see a few others leave.
The IBM deal was huge, but in all he was a disaster at GlobalFoundries surrounded by his personal sea of incompetent sycophants. Not that GF was easy, but the last thing Intel needs is a new set of incompetents in top leadership positions.
Foundry may fail, but products is facing extinction. The world has moved beyond x86 so product is playing in a shrinking market. Absent inventing and then producing what comes after ARM, products is nearing the end. If Intel is to survive, it will be as a Foundry not a Products company.
The product side is in a slow death spiral and the only piece of Intel that can survive long term is Foundry. Apart from improved versions with each gen, there has been little real product invention since the 1970's. Intel could lead vs. AMD due to better process technology, that is reversed...
Accept that the world has moved beyond x86. The next decade or so Intel and AMD will be fighting over a shrinking pie, AMD may have a future because of GPU. Intel needs a product exit strategy and then also a funding strategy to complete the bridge to foundry. I'd guess another $50B to cover...
To the original question there are 2 key factors in determining the size of a fab. 1-operational labor cost, 2-technology node. In high cost markets such as the US it takes nearly 1M wpy to largely offset the cost of operation labor vs. the capital investment. One can never close the gap to...