Could Apple Be Waitig for Mobile Fermi from Nvidia

[gadgetfreaky]

I read that the desktop versions of that won't be shipping until March in big volumes. But I have not read anything about a mobile version coming.

If there is a mobile version coming, does't that imply they may be waiting for that rather than using Nvidia. i.e. wait for both new Intel chips and New Nvidia chips.

If that's true--that implies a March--Late refresh. Don't ya think?

i.e. we might want to stop these daily--omg it's imminent threads.

 

[Sneakz]

Nvidia has no plans for mobile Fermi in the immediate future. Like not at all for 2010.

 

[niuniu]

Maybe Apple have too much stock left on th eMBPs and want to hold out another month

 

[gadgetfreaky]

Nvidia has no plans for mobile Fermi in the immediate future. Like not at all for 2010.

So no mobile cards to go against at 5850 etc? what about dx 11 support? nothing?

If they have optimus prime or whatever, why can't they come out with gpus now?

 

[gadgetfreaky]

Maybe Apple have too much stock left on th eMBPs and want to hold out another month

well dude. I have a 2007 mbp now with windows that's about to expire on it. I need to buy a mbp pronto and I don't think they are gonna refresh soon. they could be waiting for some chip from nvidia and who the hell knows when that's gonna be

 

[Sneakz]

Nvidia plans on taking on ATI's best with the GTX 285M for most of 2010 I'd assume. While it not DX11, it'll hold it own against ATI's offering. The Mobility Radeon 58XX are Juniper 5700 series based cards, and not real 5800s.

 

[gfiz]

Nvidia's got to work on actually getting a NON-mobile Fermi out before they start worrying about a notebook line....you're talking 2011+

 

[Cali3350]

Nvidia isnt planning to launch mobile Fermi based cards until April, and thats simply too long for Apple to wait in updating its line. Besides, ATI already offers fully DX11 compliant parts with very low power draw, Apple has no reason to wait for Nvidia. Optimus isnt compatible with OSX anyway (Apple and/or Nvidia would need to put in thousands of engineering hours updating its application database, something they simply wont do).

 

[Eidorian]

I'm still waiting for nVidia to get past 1.7%, woodscrews, and 280W.

 

[iKay]

It would be great if Apple whacked in a 9400M integrated and a GTX285M. That would be a serious refresh!

 

[gadgetfreaky]

Nvidia isnt planning to launch mobile Fermi based cards until April, and thats simply too long for Apple to wait in updating its line. Besides, ATI already offers fully DX11 compliant parts with very low power draw, Apple has no reason to wait for Nvidia. Optimus isnt compatible with OSX anyway (Apple and/or Nvidia would need to put in thousands of engineering hours updating its application database, something they simply wont do).

I had not heard that. April would be a LOOONG time to wait.
One would think they'd do a 330m now and then update it in april/may. But if that were true why not just do that now. Something is holding them back--all this bs about tablet timing is nonsese. If they could put it out now they would.

 

[iKay]

Well it's only been just over a year since the 17 inch MBP was refreshed to unibody and made available for pre-order. And less than a year since the first unibody 17 inch shipped.

Maybe that plays into it somewhat?

 

[deconstruct60]

Nvidia isnt planning to launch mobile Fermi based cards until April,

Seems unlikely that the newest bleeding edge performance Nvidia solution would fit the thermal constraint profile that would need to meet. The current solution is around 28W TDP. Going to come in at or under that?

Optimus isnt compatible with OSX anyway (Apple and/or Nvidia would need to put in thousands of engineering hours updating its application database, something they simply wont do).

Not just the database but new drivers. Similarly, if Apple makes Quicktime do hardware decoding on the Arrandale IGP then the database encoding is "wrong" for some apps (even if there is an equivalent windows entry). The mix of applications not offloaded from the IGP is going to change substantively from the Arrandale IGP from the much slower performing intel IGP solutions that are already there.

Also whatever "control panel" they want to come up with so that perhaps folks can "tell" the new driver subsystem which apps , or not, to send to the discrete chip for processing.

That said it would be a situation where Apple could add value. Primarily, it would be for those who run the MBPros with discrete GPUs who have the mode turned on 90% of the time. More of a coin flip for those who run with the 9400M turned on most of the time.

 

[dusk007]

Hahaha
some people really have no idea about Nvidia and the GPU development cycle.
The frist GPU based on the Gefore 2X0 Architekture for Desktop use came in summer 2008. 1,5 years later in late 2009 the first 2X0 mobile GPUs which aren't even true decedents of the Desktop 200 found their way into notebooks.
Although that is extremely slow you shouldn't expect a mobile Fermi within at least half a year after the desktop launch. Late 2010 is the soonest we will see this.

Optimus on the otherhand is a almost entirely software based technology that has absolutely Zero to do with bringing out a new mobile GPU Architecture. But is very useful for the average Notebook users (easy) and for Manufactures (easy to implement) and will probably bring a lot more Notebooks with switchable graphics and better battery life to the market.
Since it is Nvidia only until AMD releases something similar (which I wouldn't expect for a few month as it is not that easy to copy), Nvidia has a serious feature that should be more interesting for the general public than DirectX11.
You have the choice of the inconvenient switching of a comparably faster and more Power efficient AMD HD5XX0 card or you take a convenient but slower Nvidia card.
Everybody that is interested in maximum power in games and is at least a little tech savvy will want AMD and deal with the little flickering and a few seconds delay to switch. While those that cannot find the button to switch from one card to the other (those people that don't know what to do if autorun doesn't work) will be better served with Nvidia Optimus.

I wouldn't be surprised if Fermi doesn't make it into Notebooks until the next process shrink 28nm and if there is a Dx11 Nvidia part it will just be some tweaked 300M architecture which is a tweaked 8x00 architecture.

The current 9600M is a 23W TDP part and I still hope they put an HD 5830 or HD5750 into the new MBP as this would give us more speed and I prefer maximum speed over a perfectly seamless switching Optimus, which would still require Apple to get there act together and get GPU switching to Windows standards.

 

[deconstruct60]

Optimus on the otherhand is a almost entirely software based technology that has absolutely Zero to do with bringing out a new mobile GPU Architecture.

"new" in the sense have to wait for some not already released chip? No. "New" in that is requires a hardware buffer copy function engine, yes.

http://www.anandtech.com/mobile/showdoc.aspx?i=3737&p=3

That unit is sitting unused in some of the most recent releases (200M ) but no one should walk away with the notion that somehow this is (or could be) going to be retrofitted back into the previous class GPU solutions.