Should we revise this strategy after cost reduction to make it sound good and be realistic again? What do you think ?
4N5Y 4 nodes in 5 years
3N6Y 3 nodes in 6 years
2N7Y 2 nodes in 7 years
1N8Y 1 node in 8 years
0N10Y 0 node in 10 years
5N4Y or 5N5Y seems to be odd according to Moore Law
3N6Y are acceptable for one node per 2 years since Moore was already dead so you can break the rule
2N7Y is too slow for one node per 3 and half years but it is ok and understandable for a sinking ship to do that
1N8Y is the time when it cannot move anymore with 20K left and start chapter 11 filing which should have been filed with 10nm problems.
The 5N4Y effort seems to be the only thing going as planned. Some claim its not really 5 nodes, etc... but that's not the point. The point is, it was Pat who made the U-turn to finally bring in EUV ASML tools (and now high NA EUV). Without EUV, we 100% could not get any further...unless we had just become fabless.
5N4Y was bogus to begin with.
1274, 10+++ was renamed Intel 7
1276 renamed Intel 4 and 3
1278: 20A and 18A
The only node that did happen under Pat was P1276; any difference between Intel 4 and 3 is too minor (cell height difference)
In the meanwhile, the charade will continue.
Lower expectation to 4N5Y after 20K cut. Cut another 20K heads next year if it still doesn't work to 3N6Y. We don't want 2N7Y because another 20K will be chopped and 1N8Y will be back to 10nm problems again.
How about 4C5Y?
C = Customers
No need. Intel will be out of business in 2 years. Good luck to the 120K people flooding the job market.