But by the time low power ARM chips are as fast as bobcat, bobcats successor (and probably its successor) will be out, which will be much faster than bobcat.
Yes, people keep doing this, its like Pottsey banging on about Mobile graphics, in 15 years they'll be as fast as a console out now, wooo.... but where will consoles be in 15 years.
I doubt very much if that will be true. I also doubt whether most people will want to pay considerably more for much less battery life and marginal gains in performance in 90% of their usage scenarios.
The high-clocked quad-core A15s are going to be really, really quick.
It simply IS true, firstly bobcat is MILES faster already, secondly the gpu in power dwarfs what mobile gpu's can do, for certain things, crap insanely limited graphics on a tiny screen with lots of cheating and on the rails games with next to no game play and prebaked effects it can ALMOST look like its doing something great, in terms of real life gpu acceleration of flash/other stuff, the bobcat gpu is decades ahead.
Bobcat is not meant to fight ARM, its that simple, it never was, it never will. I've already said, the roadmaps are out, bobcat is moving UP not down.
As for what MS can make them do, its not MS, its owners.
You buy a phone, that you all but get free with a contract, people WANT to upgrade them every 2 years because of the increase in power in the mobile market(largely because it was untapped, its the same as building a 486dx computer a few process nodes ago and moving forwards over a few processes, it was an untapped market going through VERY quick acceleration, it WILL stop, mobiles can't get as big as full on cpu's, they can't use 40W, most importantly, they don't need to.
Anyway this is the problem for AMD, if it gets crazy fast, and enters low end laptops and even desktop THAT will dictate long term support for hardware not MS. Why because people will buy a £200 net book, or £50 on top of their phone contract every couple years, but a £400 laptop of £800 desktop, they'll keep for years, and when Android 8.6 is out, and their £500 arm PC is still going strong for e-mail and other rubbish, they'll wonder why all their apps and all their stuff is stuck on Android 5.1.
Its the market, they are in essentially a close market at the low end, closed standards, make a new phone with a new OS, no backwards compatibility is needed, the requirement to upgrade, non existant. Enter the laptop market, enter Dell wanting to sell one computer with a cheap arm and low end gaming gpu, and another computer with twice the core count and a high end gpu........ you need upgradability, a platform, you need an OS to support different hardware, you need drivers, you need everything Intel/AMD provide now for ARM to enter the "real" laptop/desktop market.
This is where overhead comes in, this is where keeping bits from old chips and not massive architecture updates comes in, and they will face the same limits as AMD/Intel, chip size, transistor counts, power, thermals.
The difference is when ARM chips launched they weren't near those limits, the first chips were essentially crap. AMD/Intel work on the limit, ARM have been no where near, they will hit it and huge generational performance jumps will stop.
Then the killer for ARM, the biggest problem. My el cheapo sendo that could text but not much else, I moved on to something that did mp3's, had a real screen, was more usable but not much more than a normal phone, then something with a camera, then a smart phone....... which does everything you could want........ what next?
Phones will hit a limit, people won't want to spend £200 more every 2 years to get a new smartphone when ultimately it does exactly nothing more than their previous phone. Netbook hit the wall, when people realised they couldn't do anything different to other devices they had, tablets will probably go the same way, phones will go the same way.
