Intel vs a whole bunch or people ARM/Broadcom/Qualcomm/Marvell/etc. It is the monolithic vs the diverse. It is a closed model vs an open model.
An analogy would be Android vs Windows Phone.
ARM IP model allows the customers greater share the revenue in the value chain and allows them greater opportunity of differentiation - much more important to un-Intel guys than squeezing last drops of performance out of a process.
ARM is much smaller than Intel as a result of course. But look at the combined revenues of the names mentioned. A powerful coalition of the unIntel-igentsia
That's how the Server landscape looked 17 years ago when the lowly Pentium Pro was released onto the Server Market. Arrayed against it was the mighty Alpha, the powerful Power, the quick PA-Risc, the snappy Sparc, the mercurial Mips. All entrenched RISCs with rich powerful backers which along with the Mainframe chips had 100% of the Server market. So, as now with Mobile, Intel started with 0% marketshare and no rich backers. A funny thing has happened over those 17 years though, Intel out executed and out priced the entrenched competition. Now Intel x86 chips have both the majority of revenue and units in the Server market. The Riscs and Mainframes are still out there but they have been pushed back into niches. Don't be surprised if a similar outcome happens in the Mobile market over the next 17 years. There will be only so long customers will accept inferior performing products from established brands until they desert you then the snowballs soon turn into avalanches of fleeing customers. Notice has been served, x86 is coming for ARM from above and below, Core and Atom in a classic pincer movement. There will be casualties ;-).
One reason why Intel succeeded is they took their overhead and amortized it over a large number of devices. Much of the engineering is reused and processor parts are fused to make new products that match the market.
Think about the words being used by Intel: CORE and UNCORE.
I imagine that Intel has a couple cores and a couple of uncores that are then improved upon and optimized for particular products in narrow markets.
Example: they probably have an UNCORE for their EX servers and an UNCORE for their EP workstations/servers and an UNCORE for their desktop clients and .....
The engineering is factored out of the industry into the Intel silicon.
Example: In the 80286 and 80386 days, Compaq Computer would get an early CPU and their engineering staff would design a superior motherboard. Being first to market they could get a premium price for their systems. Intel began to put the glue logic onto a chipset and soon, the Compaq engineering staff could not beat the easily constructed Intel reference design by enough to warrant a premium. Compaq laid off or redeployed these engineers.
Intel has accumulated sufficient IP to integrate the entire phone.
Intel is able to amortise their engineering resources over a much larger number of chips and their overhead per device is low.
The ARM market fragmentation is one of its strengths but it is also one of its weaknesses.
Vanilla ARM is no match for Intel.
ARM that is optimized by any company will have to amortise that extra cost over the number of devices manufactured.
New Intel fabs have the same effect on early devices. Intel charges the fab cost to those devices until the fab is written off and then the "cost" drops.
It is interesting that you used the analogy of Android vs. Windows Phone, when a better analogy would be android vs Apple and iOS. Like Apple's superior products, Intel will be equally successful in the mobile space by offering superior CPUs and SoCs.