Adopt ARM and do it better that they do.
7 replies (most recent on top)
@J5grzeI-pee LOL. X-Ray, don't post if you don't know what you are talking about!!
WTF, chipworks and our on competitive analysis teams use SEM and TEM to do the reverse engineering.
There is more to density than minimum gate -contact pitch you hear Mark Bohr brag about. There is minimum metal pitch, and more important things like Via placement and metal transition and many other rules that limit real density.
If Intel transistors and density was so great everyone would be at Intel's door especially Nvidia and Apple who prize performance and density, but ICF could not get any customers and has to buy their biggest one.
Go study some more before you post more B$ here!
Intel process design rules are terrible and make it almost impossible for any modern and efficient design team to use. If you got a stupidly large and monoplolstic architecture like x86 and an idiot AMD than it is easy to look good. Even than AMD chips were smalle and faster on older process in their hey day says a lot!
TMG has manipulated the product teams because they can't go outside, given a chance Intel design all would likely go to TSMC if they could, even the CPU design teams and now that Murthy leads them Sohail better watch out as Murthy is bigger and far smarter than him, LOL
@pee like customers buy chips based on density....ooohh people at Intel are blind.
You're joking.
TSMC has all the customer connections, all the relationships with all its existing customers.
ICF has what? Altera?
Talk about something you know about.
@gbm we actually do have quite a large lead in density, but other companies are turning to marketing tricks to appear to be caught up. Go to chipworks and look at the measurements they've taken with xray machines. The numbers don't lie.
Too late, Intel no longer has process lead.
And pulled out to early. Adopting RISC would've been cheaper than trying to make x86 power efficient.
We tried with StrongArm, back when Intel had a process node lead.