Yes, Intel is looking for solution. How can we help for providing ideas to Intel to compete with ARM? Can Intel take over MIPS to get entry into mobile market? Can they start with new design concept and bit ARM? In recent time, Intel bought FPGA vendor, will this help to them?
It's not just Intel vs Arm, it's also Microsoft vs. Linux. Intel/ARM and Microsoft/Linux are nearly identical competitions. ARM/Linux had their genesis and legacy on smaller systems, compact and efficient solutions, and providing more modular and configurable architectures. They are compact/efficient/nimble and easily adaptable to a variety of system scales and architectures. The current state of technology could be viewed as portable/mobile platforms having grown into ARM/Linux systems - mobile platforms and ARM/Linux have arrived at a synergistic convergence in the compute power available/needed and features sets supported in mobile applications.
On the other hand we have Intel/MS. Here we have massive platform solutions that have been grown/cultivated to fulfill desktop, server, and ultra high-end multimedia/gaming platforms. Do to the (relatively) narrow focus of development of these systems (all out massively powerful and capable systems) Intel/MS platforms are massively complex, highly integrated, inflexible, and unscaleable. Now they are trying to scale down these massively complex systems into a platform suitable to mobile computing platforms. Intel/MS must excise massive portions of their platforms that were never intended to be absent, and do so without bringing down to whole thing like a house of cards. It's like trying to take the biggest, baddest, most feature laden luxury car ever built and trying to pare it down to an efficient and effective compact car. Most people would simply opt for a car that already works well as an efficient compact car. And that's where ARM/Linux is already positioned. That's not to say Intel/MS can't get there but one has to ask at what cost and will the end product really be any good or even any better than what is already available from the ARM/Linux platform? Or will it just be more of a Frankenstein monster? But then an 800 pound gorilla behind a clunky product can sometimes trump even the most elegant of designs.
Sometime in the last year or so, I recall a quote from a senior exec at ARM, who acknowleged that eventually Intel will catch up to ARM on low power performance.
Some might wonder, when that day arrives, what incentive a handset or tablet maker would have to use Intel instead of well-entrenched ARM? If you look at Intel's recent acquisitions and their focus on SoC's, combined with their CMOS process leadership, it looks like Intel is assembling all the pieces necessary to be the price & performance leader in mobile...not today, but someday in the not-too-distant future.
Don't write off mighty Intel in the mobile market just yet!
Despite not so many positive comments about Intel (including some of my own) it is still the company to fear. Billions of dollars in revenue, still #1 (Samsung might catch up), very impressive in manufacturing and marketing. Just beat profit expectation in last earnings release. But every empire eventually falls, the questions is when and why. Might not be in our lifetime though ;-)...Kris
You have little faith in the ability of programmers at the likes of Microsoft and Adobe to continue to consume more CPU cycles for the same functionality. For instance, Word on my current machine is every bit as sluggish as the word processor on the very first Mac with greater than 2 order of magnitude difference in processor speed. Never discount Moore's corollary of software bloat.
to @yalanand: I don't think better process (22nm vs 32nm) is an issue here, yes,they are some gains in economy but the larger issue is that ARM's design is simply more power efficient than Intel's...Kris
I am kind of confused with this inference which says "Intel, however, is leading in some respects".
First paragraph says Intel is moving to 22-nm but the immediate paragraph says Medfield mobile processor will integrate the I/O with the processor and will be implemented in 32-nm
What process advantage we are talking about ?