9800M GT/8800M GTX vs 9800 gt (desktop)

Hey guys I was just @ future shop and they had this sick computer with a 30" monitor (1080p) and 8gb of ram + 9800 gt... I played gears of war on it and it was AMAZING. I was wondering how it compares with it's laptop counterpart (9800m gt), which appearantly is almost identical to the 8800M gtx

I was trying to find benchmarks but it's just too hard to find someone that compares desktop card vs a laptop card...

Thanks

 

the 8800M GTX is replaced with the 9800M GT.

They are the same specs and speeds.

The 9800M GTX is the only new high-end card that will perform 15-20% faster than the GT.

 

A 9800GT is the same as a 8800GT expect it has support for Tri-SLI. (Desktop)

My system in my signature below scores a solid 13,358 points in 3DMark06 with a eVGA 8800GT:
http://service.futuremark.com/compare?3dm06=7285602

Not sure what a 9800M GT or 8800M GTX scores, but if someone has one you can compare it to my 8800GT desktop card.

 

I know... I was asking to compare a desktop card with a laptop card. Thanks anyway (9800 GT vs 9800M GT)

 

The 9800m gt is probably about 30-40% slower than a 9800 gt and it will just become a even larger gap now that the 9800m gt 55nm has been released allowing for a further clock increase. Even the 9800m gtx will be at least 30% slower.

 

The Desktop card will eat it for breakfast. The 9800m GT is similar to a 9600 GS desktop card.

Though I think that card compares favorably to the lower end 8800 desktop card.

 

That is way off.

The 9800M GT/8800M GTX is still a high-end (256-bit mem interface) as compared to any desktop card thats semi-high (with 192-bit mem interface and under).

The 9800M GTX will perform like the desktop G92 cards from specifications... which is still freakin good for a notebook environment.

 

I've seen gaming benchmarks (of the non-synthetic variety) that suggest otherwise. I'll dig them up sometime tomorrow.

 

If its from notebookcheck don't bother

 

i dunno but ive been playin crysis in dx9 mode all high at 1680x1050 on my 9800m gt and it runs it beautifully.... so i dunno how that compares to desktops but ill freakin take it

 

So you won't mind posting screen shots with the frame rate.

 

My bet he is on 25-30 at best

 

it doesnt make any difference if it runs nicely does it.... runs better than my desktop with a 8800gt

 

^^ yup. The 9800M GT runs with a G92... like a desktop 8800 G92.

 

thats good to hear. re enforce that 9800m gt/8800m gtx performs like a desktop 8800gts. =)

 

So I guess my next question would be how does the 8800gts perform compared to the 9800GT (desktop)

 

>.> not sure why you ask that, if a 9800m gt =8800gts 312mb and and a 9800m gt is weaker than a 9800gt you already have your answer. It gets beaten by atleast30-40%.

 

If you are asking how does a 8800GTS desktop compare to a mobile 9800gt then they are pretty much the same performance if we are talking about the 640mb version.

 

The question I asked ~3 posts ago? Since we've established that 9800M gt (mobile) = 8800GTS (desktop), and that my original question was how does the mobile 9800gt compare with the desktop counterpart, my new question was how does the 8800GTS (desktop) compare with the 9800GT (desktop). Then I could know how the 9800mgt compares with the 9800gt.

EDIT: when I say 8800GTS I mean 8800GTS 640mb

 

need to be specific on what 8800gts you want it compared, because there are 3 versions.

 

I guess the 640mb one?

 

the 9800M GTX will beat the G80 GTS... but not the G92 GTS.

 

What about the 9800M GT?

 

More less the same. 9800M GT might be little bit faster. 9800M GT will never match the speed of 9800 GT because of the diffrence in SP (96 v 112) and clocks (500 1250 800 v 600 1500 900), same like 8800M GTX v 8800 GT. If you OC mobile GPU to desktop clocks than you are almost there.

 

From what I can tell, the 8800M GTX and 9800M GTX perform about the same. I am running the 8800M GTX in SLi on my Dell XPS M1730. It performs quite well.

In 3DMark06 I got a benchmark of 11536 at 1280 x 800.
In 3DMark01 I got a benchmark of 33520 at 1280 x 800.

I am running stock clock speeds using the 794.8 (7.15.11.7948) driver dated 01-20-09.

http://www.flickr.com/photos/69266103@N00/

___________________________
Dell XPS M1730
Core2 Duo T9500 2.60GHz
4GB DDR2 (2 x 2GB SODIMM)
nVidia 8800M GTX 512MB (x2) in SLi
nVidia PhysX Processor
Toshiba 320GB SATAII HDD (x2)
Panasonic BD-RE UJ-220 Blu-Ray Burner
Vista Ultimate x64/Ubuntu Hardy Heron

 

Gotta agree there but makes me wonder how powerful desktop cards are and if that level of graphical horse power is even necessary this generation which is being driven by consoles

 

yes. let the numbers speak for themselves.

 

No way to compare a desktop and laptop gpu.

 

Well the GTX 280m is pretty much a 9800m GT or GTX in terms of specs. But we will have to w8 till the sager is released with the card to be able to tell.

 

9800M GT/9800M GTX < 9800 GT
260M = 9800 GT
280M > 9800 GT

Close enough?

 

yep. 9800gtx/8800gts512>280m>9800gt

 

And 9800 GT > 260m, due to greater clock.

 

The 260M has a higher memory bandwidth than the 9800 GT though, and that's what Nvidia cards depend on; I doubt it won't perform on the same level, but we'll see soon enough.

 

Even though 260 has higher mem bandwidth but it has low cooling than desktops so there will be slight decrease in performance .

 

Wrong. 9800M GT is as fast as 8800 GS, while the 9800M GTX is a bit slower then deskop 8800 GT (in other words, 9800 GT). http://www.notebookcheck.net/Mobile-Graphics-Cards-Benchmark-List.844.0.html.

And please don't bash this source, it's allot better then your "estimations".

PS. GTX 260M = 9800 GT.

 

Note that brute power of a hardware does not equate to its performance. Laptop GPUs suffer from compacting and sometimes heat issues which desktop GPUs might not. In the same way, desktop GPUs haev other issues laptop GPUs do not have so crunching numbers is not always the most adequate way of comparing them.

On paper, a laptop GPU may be equivalent to a desktop GPU in term of pure brute power, but in real life experiences, perhaps you'll get slightly different results. There are also different drivers and other factors to include.

I know that the 9800M GS in my friend's Asus, which is supposed to compare to a desktop 9600GT, was slower than the 9600GT I had in my old desktop on a same few tests such as DMC4 and Crysis performance tests.

There is also the fact that desktop GPUs have better overclocking ability than laptop GPUs since you have less heat issues to worry about so the potential for performance might also vary here.

 

if you look at the clock speeds, the 9600gt is much faster than the 9800m gs and also faster than 9800m gts.

true, i have seen some insane overclocks on 9800gtx+, shader speed over 2ghz! core clock in the upper 800s!

 

Yeah, but I was replying to the guy who was quoting notebookcheck and they say that the 9800M GS can perform slightly less than a desktop 9600GT.

Well quite a few desktop GPUs have very good overclocking ability depending on which ones we're talking about. My old 9600GT overclocked very well actually.

Notebook GPU overclocking is always a fickle thing as you're not only stressing the part, but the laptop as a whole with the produced heat, which is more prevalent in laptops than in desktops.

 

I have a question, other than clocks, stream processers, bandwidth, can a smaller chip help performance ? like say 55nm over 65nm.

 

Yes it does. Smaller die impacts the heat and power inputs/outputs IIRC. The whole chip basically gets far more efficient with smaller chip architecture. This also improves overclocking ability obviously.

 

Shrinking the die size has no performance benefit. Terminus said aside from clocks, SP, bandwidth... so if everything else is the same then why would a smaller physical size make a difference... It will however produce less heat and consume less power, but that does not directly mean it will boost performance. It just means that for the same performance, you get a cooler and less power hungry part. Obviously, GPU companies can then make things faster by supplying more power to it; however, overclocking is more a function of the laptop's cooling ability.

 

Well if you look at it that way, I guess you can say it doesn't increase the chip's brute power as the chip itself isn't changing, but overall the chip gets more efficient which in itself is a good thing.

Overclocking ability is part of the change in architecture and allows for a same chip to be boosted higher without fear of heating issues.

 

bah..this classes of graphic cards is too close..why dont both nvidia and ati just separate gpu in just 4 or 5 classes e.g. low end, medium, high end, extreme,..then there'll be no lengthy discussion anymore..

 

Because it's not a consistent system. Today's "high end" might become tomorrow's "medium" and so on and so forth so making a classing system so basic would serve little purpose unless someone actually got hired to bother updating it every time something new came out.

 

does that mean the GTX 280m and 260m on the 55nm chip will be able to OC more than the 9800m series?

 

I believe Nvidia will be releasing and re-branding some of the 9xxx series GPUs in the new architecture as well.

Mind you, the GTX M series will be more powerful than the 9xxx series and therefore will produce more heat to begin with so OC it might not be a good idea.

 

o ok, thanks.

Do you think I can compare the laptop setup in my sig with this desktop benchmark of the 9800 GTX?

http://www.techspot.com/review/92-asus-geforce-9800-gtx/page5.html

if so...I might be able to play Crysis at decent resolutions!

 

Sorry, but I mentioned neither 9600 GT nor 9800M GS. You misred it to 9800M GT = 8800 GS.

 

sort of, the desktop 9800gtx: 675/1688/2200
vs the 280m 562/1463/950.
i guess the 280m is about 10-15% slower, so they are in the same ball park.

 

I didn't misread, you posted notebookcheck as your source and THEY compare a 9800M GS to a 9600GT desktop(read the last line of the text in this link). It had nothing to do with the info you posted, but rather the information from that site as not being 100% accurate sometimes since notebookcheck bases stuff mostly off on-paper things sometimes and doesn't take actual real life factors into account.

Therefore, I merely said that one of their "laptop to desktop" comparisons as being not totally accurate therefore I would also take their other comparisons(yours included) with a grain of salt.