It would be nice to see a side-by-side comparison of this chart from this to the one from 5 years ago. Such a graph would clearly show that even though the overall drop in CPU/server chips is falling when compared to tablets and smartphones, the change does not equate to a zero sum game. Obviously the tablets and smartphones have found a lot of new markets / customers but CPU/server chips slow down could have also come largely from slower adoption cycles.
I wonder if Apple has ever considered taking an ipad and attaching a keyboard cover, sort of like a Microsoft surface product except with an IOS OS - or maybe even an ultralight clamshell. They could try to navigate the middle ground between a MacBook Air and an iPAD in terms of cost, size, weight, capability, etc. Something with about an 11 inch screen perhaps. It could be a product type that might start to make inroads for ARM in competeing with wintel devices at the bottom end in terms of performance.
ARM promotes the mass benefit of software standards. Demonstrates the values of hardware IP standards. Yet disregards, in fact rejects the competitive advantage of commodity producer platform standards.
For mobile stronghold on first mover advantage, mass and leverage, enablement has, mass fabrication can and does, pin down the design producer base on continuous leadership to deliver a series of valued, and from time to time, uniquely beneficial offerings a handful of which achieve production scale riding high velocity proprietary board vehicles to market.
Verse Intel then, and x86 generally, ARM is a non competitor lacking chutzpah that is the tenacity to compete with Intel head to head. Not in architectural design but in production, marketing, commodity platform development, brand management and oversight.
One question does God ARM forbid it?
ARM professes but shows little coordination to win for its compliment. Disdains too rally for competitive advantage from the base of any market. Rarely runs at Intel speed and refuses to compete in Intel time. Until ARM fosters the gusto to market around and out produce Intel network ARM cluster will be contained by them.
ARM does punish, internally and externally, those who dare reconfigure for competitive advantage using Intel's own methods. Too break free from the fray of Intel pack, to punch through Intel obstacles that have always meant less than a parity position.
Can ARM cluster wolf pack?
Until ARM learns to run at Intel speed, lead in Intel time, coordinate to use Intel methods against Intel, rally foundation market and design production for Intel scale, managing to mitigate the destructive tendencies of duopoly, ARM will in fact be self Ltd.
Mike Bruzzone, Camp Marketing
Refuses to to use Intel methods that puts the cluster down and out.
and does in punish those internally, and externally, who attempt to do so.
However, there is a lot of presumed compatibity that people presume come with laptop-shaped computer, which Chromebook does not provide.
So people may be disappointed that everything they might want to add after the fact usually comes in Win or Mac flavors and may not be easily added to Chromebook, which basically wants everything to be done in a Google browser interface.
An acquired taste although I am sure some people will swear by it.
Understood Peter, but it does show how little cross-fertilization has gone on between these classes of devices. I was browsing through a Best Buy flyer this morning and saw the Chromebook. It sure looks like a low-end laptop, but the specs read more like a tablet with a keyboard. Is this really a new device category or an early sign of the redefinition of one class or the other? Maybe it is just an oddity that will fade away, but I have been very tempted to point some people towards something like that instead of a new Win8 laptop to replace XP devices.
The tablets and smartphone are extensively in use these days because of being light weight, portable and just having all features that mid manager to senior management would need. If the device sales increases definitely production volumes of these process will rise. Desktop PCs and laptops may get stagnant
It seems a little odd to say that mobile CPU penetration into the PC space is so low since much of the tablet growth is at the expense of laptops. There is some overlap, given the push for ARM cpus in the data center, but the real story is that people (myself included) are using tablets for stuff that they used to do with laptops. A laptop with a mobile CPU is essentially a tablet with a keyboard.
I don't think that's what the article said, Dylan:
"The forecast value for processors for embedded applications grew from 2 percent to 11 percent, while the forecast for application processors in mobile phones remains at 26 percent."
Sounds to me like the smartphone growth has started to flatten as well. Like all growth curves, people have a way of over-hyping the long term outlooks based on yesterday's or today's slope of the curve. The most typical shape is a S, however. Looks like embedded applications are the big winner. That makes sense, when one considers how many appliances of all types are becoming "smart."
One important finding here was that at least so far, these mobile processors are NOT finding their way into PCs. (Witness even the relative success of Surface RT vs Surface Pro tablets.) My thinking is that this might change in the future, as we get these multiple-core processors for mobile devices. But the trend is down along those lines, this past year.
Anyone looking for power, though, needs to look at x86 architecture, these days. No matter what the hype claims.
Blog Doing Math in FPGAs Tom Burke 18 comments For a recent project, I explored doing "real" (that is, non-integer) math on a Spartan 3 FPGA. FPGAs, by their nature, do integer math. That is, there's no floating-point ...