Replies to post #131053 on Intel Corp (INTC)

[wthdik2]

No other company playing in servers today can offer that advantage, and I doubt Samsung's micro-server attempts will be a serious problem for Intel (process disadvantage, x86 incumbency, etc.).

again..just as a casual observer..what I'm seeing/hearing is alot of noise from the big server customers (e.g. Facebook) about wanting some competition in the space and alot of the ARM guys deciding to take the plunge.

but...if you guys are comfortable with your moat no need to worry...

unfortunately I need to start paying attention because QCOM is one of those who have decided to take the plunge..

[Andy Grave]

.....so....for comparison's sake, let's hear your estimation of what that perf/watt metric would look like on TSMC 28nm HPM process and 20nm TSMC HPM.....I say the AMCC chip reaches about 3 on 28nm TSMC HPM vs 3.46 for Intel on 22nm..........I say AMCC chip would lead on 20nm TSMC and Intel would have to go to 14nm to counter..........I wonder what the comparison would look like with a full custom 64bit design by Apple or QCOM on 16nm+ TSMC FinFET.......I'd be worried about QCOM targeting that space at 16nm+ FinFET. How much of the server market can be served by such <$100 chips?