You Will Have a PS3 In Your Pocket In 3 Years

[zanzabar]

how else would amd keep up while being 10% or so slower in int they are better at floats (there are benches opti v xeon in raw flops for float and int), and when was the last time a game was coded that used 64bit, and not just for the sake of having it in 64bit. the only time its used in databases, and if u look amd is better at most general server things not bulk workloads of int but processing data

[masterg]

I don't doubt we could have a chip that could decode at 720p for mobile devices in 3 years. But the ability to generate video from (eg) a video game is a different than just decoding media, and I suspect that's a misleading analogy that they're taking advantage of.

the ARM cortex A8 or OMAP3530 has 720p capabilities. this is the chip that currently powers the iphone 3gs and Android. you can buy a Beagleboard with this chip with linux shell on it. not only does it have 720p capabilities, it has 3d graphics acceleration and a package on package chip that has 2gigs of ram; all on a package 12mm x 12mm running at ~.9W

as for the battery with the chip running at such low power you dont need much, and it could be supplied by lithium polymers which can get small and put out heavy doses (yet quite difficult to use)

[iTravis]

Hm...somehow I wonder does this really surprise anyone as we've seen some mobile devices with 720p outputs in the market already (like ZuneHD?) and I saw some cool gadgets in Japan that has the same capability too. So it's just a matter of time till we have a chip that is powerful enough to compare to the PS3 as everything just getting smaller and smaller.

[ryboto]

With the market moving more and more to touch screens, I don't see myself wanting to game much on a phone beyond the occasional solitaire or sudoku. If the screen is where the action is taking place, and where your hands need to go, you're looking at a clumsy experience. Maybe it would work better with phones with mechanical buttons, or if they bundled it with a small BT controller?

[ryboto]

Hm...somehow I wonder does this really surprise anyone as we've seen some mobile devices with 720p outputs in the market already (like ZuneHD?) and I saw some cool gadgets in Japan that has the same capability too. So it's just a matter of time till we have a chip that is powerful enough to compare to the PS3 as everything just getting smaller and smaller.

My phone records in 720p. It's semi-broken, since it was never updated and doesn't do 24fps perfectly, but it's 720p, and it can stream HD over DLNA. So, the claims aren't ridiculous at all.

[Hornet331]

but batteries improve...

Not really, most lithium batteries have been crawling around 100-150wh/kg for ages.

I really hope they get the Li-S technologie rolling. If they get the problems worked out, we can expect a 4 time increase of the energy/weight ratio.

[Dimitriman]

Just slap a Llano on your current psp board and there you go

PS3 graphics..

with 90w consumption

[AndrewZorn]

they said the same thing about the PS2 -> PSP
which has pretty good graphics but no not quite PS2

[Serra]

the ARM cortex A8 or OMAP3530 has 720p capabilities. this is the chip that currently powers the iphone 3gs and Android. you can buy a Beagleboard with this chip with linux shell on it. not only does it have 720p capabilities, it has 3d graphics acceleration and a package on package chip that has 2gigs of ram; all on a package 12mm x 12mm running at ~.9W

as for the battery with the chip running at such low power you dont need much, and it could be supplied by lithium polymers which can get small and put out heavy doses (yet quite difficult to use)

Sure, but adding a fully functional cell processor equivalent on top of that in a cell phone form factor in 3 years... that's a stretch. PS3 had GPU + powerful CPU. I think we'll be able to do some of the things the PS3 can, but saying it will be equally powerful is, IMO, misleading because it seems quite unlikely. I mean, IBM doesn't have plans to make anything as powerful as the cell in the next 3 years with that kind of power usage, so I don't see how this other company with a fraction of IBMs budget is going to do that PLUS more.

[Manicdan]

^ which is why this is sounding alot like a marketing gimick. sure some things can have the gpu power, but the cpu side of it sure wont also meet the 5W requirement. thats why i think if we do see a pocket gaming system in 3 years, it will not output 720p/1080p with the same power as the ps3. but 5 years id believe it. i also think that its very easy to make something look like its as strong, but really is no where near. like with the psp, some things do seem to be as good as a ps2, probably has to do with the fact its running half the resolution of the ps2, and your viewing distance is much less. id say if a benchmark was made, it would prove that its probably 1/5th the strength of the ps2, maximum.

[TJ TRICHEESE]

Gizmodo spoke with Imagination Technologies—maker of the PowerVR chip that powers smartphones like the iPhone, Droid and many others—and they said, definitively, that you'll have graphics comparable to the PlayStation 3 in 3 years.]

which is why the iphone 3Gs is the fastest smartphone on the market with its rapid 600mhz... Oh wait what about all those HTC's with the 1ghz snapdragon cpu's, we may have graphics comparable to the ps3 in 3 years but it won't be from PowerVR

[>HyperlogiK<]

Sure, but adding a fully functional cell processor equivalent on top of that in a cell phone form factor in 3 years... that's a stretch. PS3 had GPU + powerful CPU. I think we'll be able to do some of the things the PS3 can, but saying it will be equally powerful is, IMO, misleading because it seems quite unlikely. I mean, IBM doesn't have plans to make anything as powerful as the cell in the next 3 years with that kind of power usage, so I don't see how this other company with a fraction of IBMs budget is going to do that PLUS more.

I don't mean to be snarkey, but please read the original article before you post. As I pointed out once already, he didn't say that there would be mobile devices with 'equal theoretical floating point performance' he said comparable gfx. There are already some pretty SFF devices that have equal or better gfx to the PS3 like the new 11" CULV gaming machine from Alienware, the M11x.

Also as far as IBM not having 'plans to make anything as powerful as the cell in the next 3 years.' Your statement implied [to me at least] that IBM had decided that Cell is 'powerful enough for now.' Actually I'm pretty sure that all of the partners [Sony, IBM, Toshiba] have pretty much given up on doing anything further with Cell, and that IBM is just concentrating on it's more traditional POWER designs.

[Serra]

I don't mean to be snarkey, but please read the original article before you post. As I pointed out once already, he didn't say that there would be mobile devices with 'equal theoretical floating point performance' he said comparable gfx. There are already some pretty SFF devices that have equal or better gfx to the PS3 like the new 11" CULV gaming machine from Alienware, the M11x.

Also as far as IBM not having 'plans to make anything as powerful as the cell in the next 3 years.' Your statement implied [to me at least] that IBM had decided that Cell is 'powerful enough for now.' Actually I'm pretty sure that all of the partners [Sony, IBM, Toshiba] have pretty much given up on doing anything further with Cell, and that IBM is just concentrating on it's more traditional POWER designs.

You're right. I had read the article but had decided they had claimed more.

As for the IBM statement, I just meant that they don't expect to make any architecture as powerful as the cell processor in a mobile size and power envelope, not that they can't make anything better than a cell. Not that it matters of course as my premise was wrong

[>HyperlogiK<]

I just meant that they don't expect to make any architecture as powerful as the cell processor in a mobile size and power envelope.

Fair enough. I can't quite figure out whether the abandonment of Cell is a sad missed opportunity or whether the whole project was a silly idea in the first place. I was leaning towards the former, but in light of the delays and problems that faced Cell, Fermi and Larrabee I can't help wondering if all of these massively parallel general purpose computing chips aren't doomed in the first place.