Even if we make a whole deal of unlikely assumptions to get this thing in a mobile by tomorrow, can the rest of the device handle complex enough environments/models for raytracing to even matter, for instance ? In fact, with anything less than complex scenes/models tracing rays would be close to pointless, why use a ridiculously calculation intensive simulation when an approximation would yield much the same result ? Even more so I suspect on a tiny screen.
Uh oh you said the magic words that make Pottsey explode, poor Pottsey.
THe fundamental flaw in his(and Rroffs)arguments over Physx were very basic, and simple and obvious. basically exactly what you said minus the small screen bit. Phsyx does nothing better, just much more accurately, but, you can't actually tell the difference because, when a box explodes from inside, the mind can't tell where each shard should be, one goes left, one goes right, neither is "right" in the minds eye, both are fine.
What was ultimately worse with Physx is, it doesn't do things more REALISTICALLY it got more accurate numbers for unrealistic simulations of events and thats even worse. Look at Mafia 2 "physx" things, the effects are wholly unrealistic, just more power intensive.
Here we apply the same argument, ultimately accurate lighting to a 0.1mb texture on a 2 inch screen that looks woeful but lighted well, its laughable.
I've seen some mention of the effective simplicity of using a low level of raytracing to pinpoint where everything should be over some very bulky and difficult to manage and generate shadowmaps and things, I've no idea how raytracing works but for a tiny tiny device I can understand the potential to simplicity and efficiency.
Then theres the other argument "it was 2 years ago and a 75Mhz card, in a proper PC".
Sorry but the key answer here would have been, it was a 2W card two years ago and today could be done in 0.2W in a tiny form factor giving acceptable framerates, as this wasn't mentioned I'm going to go ahead and assume they can't.
As for what it actually matters, yes unfortunately dev's like anyone else insist on involving themselves in every last form factor to the detriment of every other form factor.
First PC's got overlooked because someone managed to get 15-30mil consoles in peoples hands and it felt a more attractive process, then you have phones and people started making rubbish games for them, and then tablets, and then better phones, then the new consoles, then PC's again, etc, etc.
People don't expect world class accuracy on a tiny screen, nor do dev's want the cost of developing to massively increase, the games for phones/mobile devices are cheap, short, crap and pointless, exactly like the audience they are designed for
Dev's don't want to learn how to code for raytracing and producing ultra quality stuff on mobiles that increases cost dramatically and increases time to market, just to produce "old" stuff on pc's and every other platform.
Now the question the thread, and several of the tech news sites should be asking is, where ELSE might powervr be pushing in the future and would that platform be somewhere to push forwards raytracing.
Theres been rumours of PowerVr trying to get something in some console, handheld maybe or, meh, who knows.
Sometimes companies buy other companies just to stop them being a threat, sometimes they want the quality engineers/coders, and most importantly lots of people buy companies and never ever use the IP they got from buying them.
Intel have bought lots of smaller companies with good intentions but ultimately went a completely different route, thats life.
Even in the slim chance they did bring raytracing to mobiles, the real question is, so the heck what. If they keep going forwards they can take over desktop 3d cards, really? despite not selling desktop 3d cards, and the absolute certainty that desktop and console gaming won't be based around people with a nice "phone stand" playing games infront of them, then no, theres literally no chance it will "take over" gaming.
Crap low res textures, crap ability to game over a long time due to power, crap resolution, crap screen size, crap control methods, but super doopa ultra realistic lighting............ yeah I don't think that will be something that takes the world by storm.
Revenue wise take over, possible, mobile gaming is already growing rapidly. People are much more willing to buy a 10mb game on their phone for £2 than buy a 10gb game for £30 on a whim while on a boring train ride to wherever they are going. That market is growing very fast without raytracing anyway, that won't have any bearing on it.
