I get the point but in the process node AMD was never really behind. It was after AMDs decline at 32nm that Intel got ahead. Nehalem and Sandybridge.
TSMC and GF stay a little behind Intel despite huge capacities and lots of production. They will get to 20nm while Intel gets to 14.
It doesn't help AMD too much, their main problem is still the architecture. If AMD had Intels 22nm they would still be behind and not really stand a chance.
In theory if AMD's fabs had something to produce they should pay for themselves. They could have made chips for others. I think Intel is allowing some other companies to use their 22nm node too.
I wish AMD would just come out with some magical new CPU generation but I don't see it happening.
Will be interesting what happens to this whole fab business once they are past 14nm. That is where things should start to get tricky. R&D costs will grow quite a lot.
I recently read something that this whole faster chips thing will most likely stop completely in not too long a time. The next half of the century it will be all about efficient algorithms.
TSMC and GF stay a little behind Intel despite huge capacities and lots of production. They will get to 20nm while Intel gets to 14.
It doesn't help AMD too much, their main problem is still the architecture. If AMD had Intels 22nm they would still be behind and not really stand a chance.
In theory if AMD's fabs had something to produce they should pay for themselves. They could have made chips for others. I think Intel is allowing some other companies to use their 22nm node too.
I wish AMD would just come out with some magical new CPU generation but I don't see it happening.
Will be interesting what happens to this whole fab business once they are past 14nm. That is where things should start to get tricky. R&D costs will grow quite a lot.
I recently read something that this whole faster chips thing will most likely stop completely in not too long a time. The next half of the century it will be all about efficient algorithms.