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!
That is 2013 data but it is also what I have heard from the foundries. Intel is pretty predictable with the Moore's Law .7 BEOL shrink. The foundries however use the Apple's Law shrink which is shrink it as much as you can every year so we can release a new iPhone in September.
The node name game is definitely in the foundries favor this time. I was actually very appreciative of the 16nm naming because I could easily tell who was designing to what fab. Now that they are all back in sync at 10nm it will be harder to tell.
Moving forward they key criteria will be TTM. If TSMC's 7nm is out earlier than Intel 10nm that is a clear industry win! I'm really looking forward to it actually. Hopefully it will light a fire under Intel to up their innovation game.
That is 2013 data but it is also what I have heard from the foundries. Intel is pretty predictable with the Moore's Law .7 BEOL shrink. The foundries however use the Apple's Law shrink which is shrink it as much as you can every year so we can release a new iPhone in September.
The node name game is definitely in the foundries favor this time. I was actually very appreciative of the 16nm naming because I could easily tell who was designing to what fab. Now that they are all back in sync at 10nm it will be harder to tell.
Moving forward they key criteria will be TTM. If TSMC's 7nm is out earlier than Intel 10nm that is a clear industry win! I'm really looking forward to it actually. Hopefully it will light a fire under Intel to up their innovation game.
Yes, Intel has been following its trend pretty close, 14nm node metal pitch is indeed 52 nm. But it's getting harder, so delays shouldn't be surprising.
>> Intel is pretty predictable with the Moore's Law .7 BEOL shrink. The foundries however use the Apple's Law shrink which is shrink it as much as you can every year so we can release a new iPhone in September.
Reading this raises the question - at the start/middle of moore's law , instead of the whole industry being synchronized to 18 months, would less synchronization and more open competition would have resulted in much faster innovation?
>> Intel is pretty predictable with the Moore's Law .7 BEOL shrink. The foundries however use the Apple's Law shrink which is shrink it as much as you can every year so we can release a new iPhone in September.
Reading this raises the question - at the start/middle of moore's law , instead of the whole industry being synchronized to 18 months, would less synchronization and more open competition would have resulted in much faster innovation?
Definitely. Samsung and Apple are really what changed the semiconductor manufacturing landscape. Prior to Apple leaving Samsung and Samsung racing to get them back, Intel and TSMC were happily dominating their perspective market segments. Now the semiconductor chip companies (QCOM, BRCM, etc... ) are competing with system vendors (Apple, etc....) so disruption has definitely hit the semiconductor ecosystem, absolutely.
