9800 MGTX soon out?

The new INF version 3.248 shows new MGPUs, hopefully they will come soon

QUOTE
NVIDIA_G92.DEV_0608.1 = "NVIDIA GeForce 9800M GTX"
NVIDIA_G94.DEV_062A.1 = "NVIDIA GeForce 9800M GS"
NVIDIA_G94.DEV_0628.1 = "NVIDIA GeForce 9800M GTS"
NVIDIA_G96.DEV_0648.1 = "NVIDIA GeForce 9600M GS"
NVIDIA_G96.DEV_064A.1 = "NVIDIA GeForce 9700M GS"
NVIDIA_G96.DEV_064B.1 = "NVIDIA GeForce 9500M G"
NVIDIA_G96.DEV_064C.1 = "NVIDIA NB9P-NS"
NVIDIA_G98.DEV_06E8.1 = "NVIDIA GeForce 9300M GS"
NVIDIA_G98.DEV_06E9.1 = "NVIDIA GeForce 9400M GS"

 

The 9800 isn't even out for desktops yet, so I doubt it...

 

don't hold yer breath. but they showed that in the INF for the 3600m then a few months later it appeared.

 

Although they`re in the inf, they won`t be available until at least summer, talking about the desktop versions...
For the mobile, wait a lot more, maybe fall 2008.

 

Mind if I ask a stupid question - that's too bad, here goes anyway - anyone suppose that maybe NVidia dumps these heretofore unknown GPU identifiers in their INFs just for funsies, to see what sort of a rise they can get out of folks? Or, at the very least, keep the buzz going by getting everyone to react in a pavlovian manner?

I dunno, I'm sometimes pretty dumb (hey, I may be dumb, but at least I ain't stupid - at least I don't think so ), but it seems to me that this is more a case of the Emperor's New Clothes than anything else.

 

What about the fact that notebook usage has picked up dramatically over the past few years. Wouldn't brands try to get mobile hardware out more quickly due to the more constant demand for it? my two cents.

 

I am sure honda civic with come with V12 engine one day

 

Mobile GPU versions are usually out 6 months after desktops.

 

Yeah..look at how long it took for the 8800m's to come out:too long.

 

:twitcy: In high-school, I had an acquaintance who tried to shoe-horn a small-block chevy V8 into a civic; it didn't work out too well.

 

yes but, why? i mean, the notebook market is growing immense, so there should n't be this form of delay like with console > pc games... crazy

 

I think it`s because of the evergrowing power requirements of the ultra high end cards(check 88M GTX- about 150W in full load) and reducing them to about 1/3 of that (45W for the 88M GTX) to fit in a powerful notebook...sounds tricky enough to me to be worth at least 6 months of work.

 

die shrink process

 

It's not that simple as you may think. Try understanding that these cards are only for enthusiasts and they cost an extra to manufacture. It's not quite like "die shrink" and they're released.

 

Fade's correct; if these were jewelry (they're almost as purty, IMHO ), these things would be Tiffany's not Zales.

Also, notwithstanding that there have been big steps made toward standardization in notebooks, they are still essentially one-of-a-kind boutique items - yes, even the mass-produced business-lease notebooks are essentially one-off items. Given that fact, even though NVidia produces one 8800M, there are, technically speaking, as many different, incompatible, 8800Ms as there are different manufacturers. That means that there has to be much more customization of each piece of hardware for the notebooks than there is for the desktops, and that additional customization adds a lot of extra time to the eventual retail release of the notebook GPUs.

 

Man,try turning your Dodge Caliber into a Smart car with the same power, but only 1/3 the consumption. Tell me when you`re done

 

I'll help with some beer .

 

And I`ll help drink it

 

Who wants a dodge, smart or not? In that category, I'll take the White Zombie any day of the week(end).

This 1972 Datsun 1200 sedan:

source: http://www.plasmaboyracing.com/graphics/95datshw.jpg

will smoke almost anything you can put up against it, all without burning a single drop of gasoline. The specs webpage gives a 0-60 time of approximately 3.5 seconds.

Here, a Chevy Nova with a "built V8" gets ready to be "absolutely wasted" by what looks like a granny special:

source: http://www.plasmaboyracing.com/graphics/grynva.jpg

Oh, the joys of being able to put all of your torque to work at once!

Okay, so that was gratuitously off-topic, except for the electronic gizmo modding part , but you gotta admit, that's pretty d**ned cool.

 

i didn't say it was simple

but that is the process.

 

Theoretically I could fly. I just put myself some wings and jump from a very high building. This is the process, but it ain't that simple.
And you're right, they can do it, but it's not very profitably for nVidia to release a new card now. They have to make some real profit from this one.

 

Actually, I think that the 88M GTX pretty much is a "block buster" , if you know what I mean. A LOT of new laptops came out with it and the fact is , actual games require something more powerful, and since SLI is still crap(to me is SLI can`t give at least 180% performance it`s crap) a new one is pushed by the need of the market.

 

Wings, shmings! Just jump and forget to hit the ground!

 

Made my heart warm and fuzzy about flying : http://www.youtube.com/watch?v=_ovk...nd/profile/displayProfile.do?userid=112435745

 

Hee-hee! God bless the Kiwis!

 

Yea, it was kind`of emotional. Something like make your dreams come true, regardless.
Now where would I find a chip to make my own 88M GTX...

 

gee i started a whole offtopic row....

y'all should fly to harry potter and let him shrink a few.


i just wondered why it isn't able to do faster as the notebookmarket is rising so assbadly.

 

Largely, I think, because the time-to-market (i.e., development time from initial NVidia prototype to final release-to-market vendor version) is a lot longer for notebook GPUs than for desktop GPUs. Assuming that the notebook version is essentially based on the desktop foundation, so it leverages off the basic architecture developed by NVidia, you're starting from the same point in time development-wise, but the desktop version will always be released earlier because there is much less customization required in order to get to the RTM vendor version. Unfortunately, that extra development time will remain until notebooks become more uniform cross-platform.

 

whats with the off topic stuff? i dont get even half of it LOL

 

OK, well to be on Topic...

Have you noticed the sudden drop in 8800 DT prices?
ie. u can get 8800gtx sli for 1500

maybe its hinting to something

 

may I point out that the 8800M GTX and the 8800 GTX are NOT equally powerfull?

 

May I point out that that is besides the point, and you were supose to read between the lines?

 

On topic

I think that the big jump on 9800M GTX will be the all 128 SPs disponible... 8800M GTX have 32 less SPs(laser cut) VS 8800 GT(96 SPs VS 128 SPs).

That will give more 1/4 of the shader processing power, or even more, if the shader processing units suffer a frequency rise.


The G92 core revision used on the 9800 GTX is an improved version of G92 8800 GT. Because 9800 GX2(like 3870 X2), the core have been improved on thermal specs, releasing less heat on load conditions.

9800M GTX will be the same as an 8800M GTX... same core(diferent revision), maybe a little higher core/shader and memory freq, maybe 128 SPs VS 96 SPs, PureVideo2...

 

I'm sorry.
the text between the lines was too small :tongue:

 

^

Audigy, my main desire would be to see a 9800MGTX x2 in the M57xRU Whatta you say,care to take a bugging trip to clevo ?

 

well only time will tell...
and where is the sli pricedrop?

 

Should there be one?

 

Has nvidia ever tried to split up the production of desktop and notebook? I wonder if this would be effecient however. By doing this though you could have two different teams starting with their own needs and goals. Then its not a matter of shrinking its a matter of where to start I guess. I dunno. maybe this is a dumb idea.

 

It could be a brilliant idea, but they don't usually follow our needs.

 

That's not a dumb idea, so don't be hard on yourself (that's the job for the rest of the posters here ). Unfortunately, I'm not so sure how well it would work out in practice given that there would still have to be a lot of underlying commonality given that the same OS and the same apps are expected to run on both laptops and desktops, in addition to the desire for efficient use of resources by leveraging one off the other.

Now, it might then make sense to start with the notebook GPU design first, and then leverage that development to work on the desktop GPU; however, the flagship GPUs are always the desktop GPUs, so that would probably be seen as putting the cart before the horse.

Ultimately, the most likely route would probably be to make NVidia's MXM wish-list into a real bona-fide standard that applied cross-platform to any ODM/OEM that wished to use NVidia GPUs in their product. That would probably cut out a lot of the hooey-balooey that currently goes into customizing the GPU for each and every ODM/OEM (and each and every different model line produced by each and every ODM/OEM), thereby significantly shortening the time-to-retail on notebook GPUs.

 

well,

yes?

 

88GTX SLi for 1500 the whole system or just the gpus?
Cause you can get the 88GTX for about 500$ each. => 1k $ for SLI

 

ya, don't ask me! lol