Differences? 8800M GTX and 9800M GT

Nothing to do with DDR3 or DDR2. That guy had no clue what he was talking about.

 

Ok, which one? to me it looks like the 9800m gt is but I cant confirm this...

 

The 9800M GT is slower than the 8800M GTX.

It looks to be compareable to an 8800M GTS. With maybe some slightly higher benchmarks due to efficiency improvements.

I have seen a review of the 860TU with 8800M GTX and T9400 (2.53) score around 9900 3DMark06.
(some re-sellers in other countries are offering the 860TU with 8800M GTX)
I guess you can expect same system with 9800M GT to score maybe 8000-8500.

 

review... where?

edit: nvm. sorry, first post i read when gettin back on these forums.

i think his talking bout this one mentioned here.

http://translate.google.com/transla...kler/multicom_kunshan_m860/53697/3&hl=en&sa=G

 

^^ Would you be able to share that review by any chance?

 

Its in Norweigan, but here you go:
http://www.hardware.no/artikler/multicom_kunshan_m860/53697/1

It has been posted in some other threads already so apologies for the multiple post.

 

Solely from a commercial perspective we may actually deduct the "standings". Please look at Eurocom's offer : http://www.eurocom.com at the "VGA Technology" dropdown box for the EUROCOM M860TU Montebello. 8800M GTX is 150$ more expensive than 9800M GT which in turn is 100$ more expensive than the 9800M GTS. There's also another weaker card offer still from Nvidia but the best bang for the buck looks to be the 3870 which may prove to be way better and is 100$ cheaper than the 9800M GTS!!!
In my view it's odd to have such a high difference of price for the clamed "statistically insignificant" difference in performance between the 8800M GTX and 9800M GT. The logical conclusion is that the 8800M GTX is better, and it remains to be seen how much better, than the 9800 GT.

 

im still betting that the 8800gtx in real world use is going to be better than the 9800gt.. im looking more into this now i know someone who ordered his sager with the 9800gt so when he gets it if no one else has posted them number he said i can have it for the day and run it through whatever i want to.. hopefully he will have it this week.. if not early next week..

 

I think we've been bamboozled by the resellers....now who's supplying the misinformation? Can't blame PCMW anymore! I didn't wait this long to have my head messed with - tomorrow i'm changing my order back to the 9800gtx or the 8800gtx.

 

Everyone is just speculating right now. You will never know for sure with out some real world tests.

 

There wouldn't be any need for speculation if the resellers would just give us the specs on the 9800m gt

 

Can this review be trusted? If it is indeed true, then the 860tu is much weaker than I expected. Also, what was that about "noise and heat"?

 

how so? 9897 in 3dmark06 is nowhere near "weak"

 

Let's wait just a minute here folks.

I find it difficult to believe that there are so many here with opinions, conjecture and speculation who also find it so easy to sling this stuff as if it was somehow provable truth.

As I have said several times already, the performance, power consumption and heat generation of both cards is essentially identical, given that they are in the same model laptop with the same configuration, and any difference detected will be statistically insignificant. It will be as statistically as insignificant as the difference you will find if you were to test two identical models with identical configurations including two 8800M GTX cards.

Now, one can site all of the technical theory you want to, but the proof is in the outcome. So, before anyone advances any more insults directed towards me or any other person here, I would suggest you wait to see what testing really proves out, because if you don't the thing about writing is that it lasts forever and it is painful to have to eat all of those "expert" words that were spoken in youthful exuberance.

Further "logical conclusions" based on pricing only shows a total lack of experience and understanding of how things really work. All you have to do is look back at previous introductions of next generation video cards to understand what I mean. Video cards, like processors, are often "bracketed" in their performance and price. So, while the 8800M GTX was current it was in the top bracket. Now the 9800M GTX occupies that bracket. The 9800M GT is in the next bracket down, but since it is the next generation card it is not unusual to expect that the 2nd bracket for the new generation = the old top bracket in performance, but of course would not = in price.

Enough said on this for now. Let's wait to see what the real world testing shows over time...

 

Paladin, again I see a good explanation but still no specs given for the 9800m gt.

 

Nvidia doesn't even have the specs up for the card, how would the resellers get it?

 

Well, I agree with what Neil had said earlier. The only reason that it scored a 9897 was because of the DRR3 ram, which boosted the score up by a large amount, which compensates for the lower GPU score compared to the 8800GTX.

I don't really care about the lower graphics card performance, but if it really does have a short battery life, loud and produces a lot of heat as said on the review, then that defeats the whole purpose of this laptop.

 

um....they are selling them so they should know them. Would you buy a car without knowing whats included?

 

So Donald just cleared the "Clearing the misconception: The 9800GT is NOT close to the 8800M GTX! Read". Nice.

there is a different between business partner and public

 

I would still like to know more information about the noise and heat produced by these things under different conditions and the battery life. How many minutes under this so called "silent mode"?

 

Neil did not say anything about the performance boost due to DDR3. Quote him if you can.

 

Nirvana,
We'll see bro. I know you want to see the scoop, we'll see what they'll say when we see both systems benched on equivalent setups.
I guarantee that the 8800M is better. DO people really think that 64 shaders versus 96 is compensated for by clockspeed? If so then how come good old GPU-Z showed that it was slower?
C'mon paladin44, out with it. I just want to know what you're telling us with a basis, I've backed up everything I said with statistics and NUMBERS.
That screenshot in GPU Z showing this carfds specs supports my thesis.

 

We need an actual review to set all this crap to rest.

 

what specs in gpu-z are you talking about all the important information is missing?

 

Why is everyone so worried about the slight difference between the 9800gt and 8800gtx. 9800gt is a lot cheaper and I wouldn't mind if it was slower than 8800gtx. Anyways, I think we should be more worried about the HEAT, possible NOISE, and BATTERY LIFE.

 

NoelGallagher,
As one of 8800m GTX owners, I will be more than happen to find out it is better. But I have some questions for you tho.

Where'd you find the clocking and number of SPs information?
also, I don't think "shader units" equals "cores"

 

http://forum.notebookreview.com/attachment.php?attachmentid=21049&d=1216151161

there are your specs.
How the hell is this GPU even CLOSE to the 8800M, keep in mind that those aren't "downclocked" since those are fresh off vantage!

 

@Paladin - While I respect what you say because you are a reseller, you really have no proof. All that you are saying is that they are identical, with no actual benchmarks or anything. A great way to compare them would be to test them under Crysis : UT3 : 3dmark06 because those programs are pretty GPU intensive (minus 3dmark because that also involves the CPU).

 

I must be blind because I can't see anything about SPs/shaders number.

 

what in the world are you talking about....the clocks are the exact same as the 8800m gtx and the shader core information is missing. You provide just as little proof as they do. Information is great....but people seem to lack the concept of proving what they say.

 

The proof is here:

Scores have been displayed in another thread you can view HERE (9800M GT) and some later OC'd Scores HERE

If you use the URL in that post, it points to Chaz's review of the 5791 HERE (8800M GTX)

The main differences between the two configurations above is the CPU (X9000 vs X9100), HDD and Vista Edition.

 

Since I'd rather buy the NP8660 with the 9800M GT over the NP5796 with 9800M GTX anytime due to portability being very important for me I'll gladly eat humble pie if what you are saying will be backed up by real in-game performance (comparable fps between similarly configured NP8660 with a 9800M GT and a Eurocom M860TU with the 8800M GTX for example). However for the moment I remain skeptical.

 

except the problem with that is the clocks provided with 64 shader cores does not= 360 Gigaflops. Not only that but if you look at the 9800 gts its uses the same amount of cores yet only has 240 Gigaflops
http://www.nvidia.com/object/geforce_9800m_gts.htm
which would mean the 9800m gts is under clocked compared to the 9800m gt

Wow I sure hope this isn't what you are talking about.....if it is, you really do not have any clue what you are talking. As a matter of fact if you actually looked, mine even has a higher texture fill rate than the specs you gave. Mine has 28.00 GT and the specs you gave say it should only have 24.0GT.....

If your so eager to prove your system, give us a pic with 3dmark gpu-z and have the current day in the picture.

 

If it's 1600 mhz effective it shows up that way when I gpuz'd my setup ages ago.

it did for me when I had my m15x. IT sure looks like 800 mhz effective.

 

I get mad when peopel think I have something to prove by lying about my material possessions over the internet.

 

@bwhxeon and NoelGallagher,
the graph is definitely invaild, because not only the clocking is wrong, but also my texture fillrate is showing "33.6GTexel/s"

 

Nice pictures neo...must have been to hard to actually turn it on though.

 

You cant compare whats online with the actual product as they are all different in some way. Check the specs on different manufacture desktop cards online sometime and you will see what I mean.

 

Okay I think its been proven that he does own a desktop. I have deleted these offensive posts and gave an appropriate infraction.

Please keep the discussions civil and factual. No need to go juvenile over a graphics card.

-flipfire

 

I was thinking the same thing.

 

You obviously think you have something to prove about your ego over the internet. I've been building computers since I was seven, I owned and operated a sales/repair business with my bro that netted a decent amount of cash for my age, and im only 19. Big whoop, that does not make me some sort of expert who has the right to shoot down all opposition to my opinion, nor does it make me the master off all things computer. All ive seen from you is that you like to boast a lot of numbers and shoot down anyone who so much as suggests a speck of information that may not agree with something you have stated.

As far as I know, forums are used for discussion, not just to post YOUR version of the gospel which you ironically defend to no end while using lines about "dont believe gospel" from other resellers. Not only that, you are going on VERY limited information from the NVidia site. From what i see on there the only factual information you have is the number of cores and gigaflops. From the GPU-Z, you dont know A LOT of information. Justin's SS does not show a lot of critical info needed in making the assertions you have made.

Lastly, don't go attacking other members of the forum; its childish, rude, and disrespectful. Paladin and bwhxeon have been thousands of times more informative and helpful to me, and many other than you have been. If you want to post pictures of yourself with rude messages on a piece of paper, get a facebook dedicated to your ego or something.

EDIT: Also, taking GPU-Z as fact is a very funny concept, as im sure a LOT of people will tell you, it constantly reports information incorrectly.

 

No, 8800M GTS is a G92 stripped down to 64 SPs while the 9800M GT is a G94. They may be identical on paper but they won't in a real world. I'm really looking forward to benchmarks. The 768MB 9600GSO (96SPs G92, 192bits) is equal to the 512MB 9600GT (64 SPs G94, 256bits bus) so...

Do we have the 9800M GT's clocks? If it has the same ones as the 9600GT, it probably will meet the 8800M GTX

 

http://www.nvidia.com/object/geforce_9800m_gt.html

Compute Cores: 64

Gigaflops: 360 (so we can suggest the clock higher than 500/1250)

 

hit me with a waterballon if this does not even come into the equation of the topic - but here goes:

could/would the 64bit OS offer any benefits to any results associated with any of these tests?
are not the recent tests based on 64 bit vista vice 32?

just curious....

(goes and gets into watersuit)

 

@ Paladin, if i recall correctly you guys got test units last week. Would you be so kind as to run some benches for us so we can put this to rest? a 9800m gt with a P8600 or P9500 processor would be great.

i completely understand where everyone is coming from because the facts are all showing that the 9800 is slower.

 

The 9800M GT seems to use G94, so it a doubled G96 (used on a 9600M GT).

If the 9800M GTS has 240 Gigaflops and the 9600M GT 120 Gigaflops = shader should have the same clock. 1250 MHz should it be.

The 9800 GT has 360 GFlops, so the the shaderclock should be 1,5 x higher. 1250 MHz x 1.5 = 1875 MHz.

So it seems the 9800 GT shaderunit is clocked far behind the desktop 9600 GT. (1650 MHz and 312 GFlops).

 

sorry - but can you actually compare book gpu's to desktop ones and remain fair about all of this?
i mean they are not even the same are they?
i would assume one uses different things over the other...
but then again, i am not a guru on this stuff...just a newb

 

Hauser, regarding the 9800m GT shader clock, maybe you should take a look at this? http://forum.notebookreview.com/attachment.php?attachmentid=21049&d=1216151161

 

The GPUs are the same. Actually nVidia has four G9x GPUs. G92, G94,G96 and G98. For mobile card nVidias makes simply then a preselection of the chips by means of power consumption and thermal cycling. Some chips runs on lower voltage and generates lesser heat. This one are used for the "mobile" versions.

For example the 8800M GTX used a stripped down G92. The same GPU you will find on the desktop 9800 GTX or 8800 GTS (512/1024 MB Version).

nVidia doesn't develop special "mobile" chips since a long time.

So if nVidia says the 9800M GT has only 64 SPs/shader cores, the only possible GPU used for this are a stripped down G92 or a G94. Since stipping down a "big" GPU is very expensive it can only be G94. The same GPU is used on the desktop 9600 GT.

 

gotcha - thx for the educational information - now more of it makes sense....