Not only is Intel's business model under pressure due to custom requirements, it is also at risk just because of ARM's existence in the server space. The lowest retail price for an Intel Xeon E5 (Ivy Bridge) is $225. The ARM-based server chip companies are coming from a world where getting $50 per chip is a huge price. (Early entrants like Calxeda can charge a lot now, but not forever...)
So even if ARM partners are not successful in the server space, their mere presence will force Intel to lower its pricing. This is bad news for Intel because it gets most if its profit from those server chips.
Should Intel start making pseudo-custom CPU's, i would be more concerned about "at what cost".
In the not too distant past, Intel has always had their R&D expenditures paid for up front by higher costs that only went down on two SCHEDULED price decreases per year when they had NO competition. Only with competition did their pricing change to a market driven price structure.
Of course, competing with ARM processors automatically makes Intel charge appropriate pricing until you start making custom cores which gives the advantage back to Intel on pricing.
Too bad we can't go more towards a PCB pricing structure where there is a direct cost per square inch at a particular fab process since it doesn't matter what the blend of parts is (within reason). But that is just a fancy way of normalizing the R&D cost and since their processes are at least one node ahead give them a license to charge a premium.
I am old enough to have lived in a world without PCs, without internet and beeing in the semiconductor industry and there being responsable for the in those days largest european player in the computing market, I have seen the changes taking pace over the decades.
I remember that once in a US superstore I asked an old couple, by now I am probably older then they were in those days, why they were purchasing PC. The answer gave me the compeeling reason why the internet would take off! they told me their kids were living all around the USA and that email was the mean by which they could stay in close touch at low costs!
Recently I have seen a factor I consider similar in its importance to the future. my kids, they all always had the latest and fanciest PCs available to them and I succeeded to have them not just play with them, but to use them as tools together with the internet. Still so, the PCs have moved to a corner in their rooms and smartphone and tablets have taken over! They simply are a post PC generation that have changed their pattern of usage. That is the compelling reason, why the future is not with traditional PCs and not with Microsoft OS.
I am not saying that PCs will fully fade away, we still have mainframes and the last player remaining in this biz, IBM, is making good profits.
Also, my usage pattern has better use of a PC with pre Wondows 8.0 OS. I do not know your experience. But nobody but myself is willing to work on my messy keyboard! Could you imagine how a screen would look like pretty soon if I was using its touch sensing features, as required by post Windows 7 OS?
I am scared to death, when Microsoft will cancel one day its support for Windows 7 like it has just done with Windows XP! Looks like then I will be a dinosaur and destiny to extinct would be much better!
Dunno about ARM, but i can buy TODAY an Avoton based server motherboard from Asrock : 4 cores, 64bits, up to 64Gb support, Mini-ITX, 14W TDP and of course passive cooling ... for 259€ >> http://www.servershop-bayern.de/index.php?page=product&info=9583
If you want to win a battle, you have to be at least ahead of your competitor ...
Those of us who are old enough call engineering appliances Applicon and Calma. Large Japanese companies built special-purpose hardware for chip routing. These solutions disappeared for a good reason, and aren't coming back.
I couldn't envision using a laptop 20 yrs ago to do my eng work. Now I plug it into my docking station hooked up to a 27" display/keyboard/mouse. All I would need to switch to a tablet/smart phone is a docking station and the apps. Would help to have a portable display/keyboard/SSD for trips; seems easy to build and be light-weight. Got the same problem w/my fingers and eyes.
About Jim. What else can you expect from a bitter disgruntled ex-employee like Jim. He has nothing but negative comments about Intel. I dont expect anything else from him anyway. ARM expects nothing more than 10% market share in the server space by 2017. Yes, 2017. Jim makes a big deal about AMD and Nvidia Server. None of these 2 are shipping anything today. At best, AMD will have something in 2015 and Nvidia, God knows when. Jim, you are defnintely living in a La La land.
Perhaps I should have been clearer on the use of the word 'traditional'. If you do engineering work you will end up on an engineering workstation not a traditional PC. If you do video editing it will be on a video editing computer. I just thik we are heading for a 'break-up' of a monolithic version of a PC (traditional) and headed for a more fragmented market. As a current example, is a 'gamer' PC a traditional PC- I think not.
NASA's Orion Flight Software Production Systems Manager Darrel G. Raines joins Planet Analog Editor Steve Taranovich and Embedded.com Editor Max Maxfield to talk about embedded flight software used on the Mars on EE Times Radio. Live radio show and live chat. Get your questions ready.