Hey Nenni, get back to bragging about ARM64 the first good news out of the ARM camp in two years. Here let me help your hyperbole "The industry can't possibly have two ISA's so they will pick ARM." Consider Intel's delay of a few months on 14nm a gift. It's a gift of time that ARMH will squander just like they have these past two years. The net-net is that ARM market share will continue to fall to Intel.
Intc has been doing finfet for two years now.There having problems with 14 nm true .Heres a question you think tsmc is just going to walk in and start doing 14/16 finfet when they have never done finfet yet?
I don't think so.
Early August at Semicon West I heard that the 14nm move in was delayed 3-6 months. It was confirmed at a semiconductor equipment analyst lunch late August. This is all second hand information but I believe it to be true.
Brian K said 14nm was not delayed at the IDF conference in early September. It is mentioned in this article:
Intel Quark: Synthesizable Core But You Can't Have It
by Paul McLellan
Published on 09-10-2013 11:43 AM
That story changed on the conference call last week. Brian sited defect density issues (yield). I asked a friend from Intel about this and here is what I was told;
Intel develops the processes in an R&D facility in OR. Once the 14nm process is stable they will replicate it in production fabs in OR, AZ and then Ireland.
So yes, in fact, equipment was delayed. The problem has not been resolved yet otherwise Brian would have clearly stated it was (my opinion).
But you are right, so what if 14nm is delayed a quarter or two. It really does not make much of a difference to Intel.
It does however give a little more breathing room for the fabless guys. Altera 14nm will be delayed so that is good for Xilinx. Intel 14nm SoCs will be delayed so that is good for Apple, QCOM, and Samsung.
My guess is that 14nm SoCs will hit the market from all sources in Q1 2015 so no, Intel does not have a 3-4 year lead for the mobile market.