I can't believe that ashraf.eassa says that no one will buy ARM based servers. Not even ARM itself - for its compute farms ? or even Nvidia ? or even Apple ? or even Qualcomm? or even Broadcom? or even Sony? or even Toshiba ? or even Canon ? or even Samsung ? or even Amazon ? or even Facebook ? or microsoft ? who else is building ARM chips has the incentive to do that , it will pull all of the EDA industry at least . that is a dent at least :)
Ehh! What's the story ? is it that difficult to think dynamically instead of static - hard coded monopolistic - we will conquer all (reminder - Napolean failing in Russia) - pattern of thinking !
The ARM business model is more naturalistic than intel's , everyone has an incentive to make it win, EVERYONE. So everyone joins forces! but intel's business model is a catastrophic one: we will conquer all!
I work in the EDA, just compare the EDA business market size before ARM became a hit and now . everyone is benefiting from Arm's success !
You apparently don't know much about the server business. ARM is only competing for a small percentage of the server business. The don't have the ecosystem and the software to compete in most of it.
And in the area they do compete Intel pretty much sealed that off with the new Avoton release.
In the meantime Intel is expanding its ecosystem into the rest of the datacenter, meaning that they have better and more comprehensive solutions that ARM is unable to match.
ARM's chances of serious penetration into the server market are just not that good. It's commonly acknowledged. Read up instead of posting nonsensical mumbo jumbo about Napoleon and how ARM's business model is more naturalistic.
And you also have to throw in ARM's serious fabrication problems. 20nm is economically damaged because it's planar and ARM won't have FinFET for years and years. You might as well read up on fabrication while you are at it. A little knowledge will help you grasp the dynamics you are currently missing.
INTC will lose those powers soon, starting with the server market. High end Xeons currently sell for +/- $3000. In the very near future (less than 2 years) for that same price you can get 30 (!!!) or more ARM-64s (with 8, 12 or even 16 cores each) with parallel computing.
ARM will penetrate the server market first, then the desktop market.