9800M-GTS beats 8800-GTX reality or bollocks talk?

This is where we stand.

[email protected] 88GTX@600/1570/970 FSB 800/667 [email protected] 98GTS@698/1798/988 FSB 800?/1033


1900x1200

19.795 17.705

+2.09


1499x1050

32.075 28.365

+3.71


1680x1050

26.525


1280x1024

35.075 31.690

+3.385

So on an even platform; I doubt very seriously that notebookcheck only had 1FPS difference with all being the same and no overclock.

 

yeah, the 7811 wont go higher without messing with the front side bus, which it cant do at this time.

 

they ran a test against an 8800m gtx??? where that test at??

 

No, against the 98GT and 98GTX.

They claimed that the 98GTX was like 2 to 3 FPS more than the GTS. And that the 98GT was like 1 FPS more, I'll post the link.

 

I'm sure with your testing, you know that isn't right. Heck, you have lower scores than E-Wrecked. LOL!

http://www.notebookcheck.com/NVIDIA-GeForce-9800M-GTS.9893.0.html

 

Actually.. My CPU is downclocked just a touch.. stock it's a 11.5 multiplier, but runs @ 11x with RMClock. Cant get the extra half. So, I'm closer to 2.9ish And run yours stock, no overclock...VS mine stock..no OC on GPU/FSB/STOCK. just the 1280x1024.

 

LOL!

You always have some sort of an excuse; it must be in the bloodline? I’m cool with what went down, this is over. Johnksss got 27FPS with overclocked GPU; mine will beat that at default. So anything else is irrelevant. Now we just need someone who will give an honest score with 98GT and 98GTX.

 

As I ve posted in the other thread. GTS, which is a civic with couple of turbos can not beat a GTX, which is a ferrari. It might keep up on the straights but in the corners the Ferrari will blast it.

 

well...looking at their little chart there... that is very high and not on high

so run yours at very high no over clock at 1280x1024 & 1600x1200

 

did you get your gtx machine back mastamarek??

 

I guess that the original purpose of this thread is now solved, the 8800m GTX will beat the 9800m GTS like it was supposed to, and that Notebookcheck either messed up something, did not run the tests on even platforms, or altered the scores on purpose.

Although I agree that there should be a Stock 8800m GTX VS Stock 9800m GTS comparation for informative purposes on this thread.

I am anxious to see how does the 8800m GTX / 9800m GT compare to the 9800m GTX.

Cheers and keep the playground fair, guys.

 

Tomorrow at 12pm, can not wait

 

about that...i only gained like 2 to 3 frames going from stock to over clocked, but i can't remember. dammn i might have to boot my drives in another machine to see if i can find those runs.

nice!!!! now im asking for a single and a dual card run, since no one else seems to want to do this. which i don't understand why...

and i think im going to be ordering mine this week. dual gt's

 

Yeah, I ll try to post as many benches as possible but what do u want me to run it on? I ll be installing Crysis and GRAW2 for now I think. Rest of my games are on the 7811.

 

im not picky. 1280x1024, 1600x1200 (1920x1200 only if you have it) high/very high. so we can get to the bottom of notebookcheck's comparison. didn't zenica run a crysis test with his 9800m gt?? im so tired of searching and looking that i gave up for a day..*LOL* i have been looking at every 9800/8800 anything thread for the past 5 days.

 

Well someone over at the sager forum claimed he got 47 fps average in DX10 1920x1200 which I really doubt. Now I don´t know if he had a tweaked Crysis or anything, however two 9800m GT in SLI isn´t that much better nor faster than two 8800m GTX in SLI. So I very much doubt those scores.

 

well..magnus, since you don't care to run stock to stock..which by the way, is how notebookcheck compares cards..not OC'd to the stable limits.. I'll go ahead and post mine. brb

 

8800m GTS and 9800m GTS is not too far away in Crysis at least at 1920x1200. 2-3 frames per average differ. Though the benchmark doesn´t involve any firefights so in real game it might differ more.

Better up is to run the benchmark on some of the heavier levels and see what kind of difference we will get.

Now if I had an extreme CPU I would probably get 2-3 more fps average at 1920x1200 High

However I do have a config now that just looks plain awesome and averages 37 fps at 1920x1200. Thread here, scroll down a page and you will see some screenshots too. I posted the config too autoexec.cfg so do try it out and do run the game in DX9. Just place autoexec.cfg directly in the Crysis root folder and be sure you have no system.cfg tweaks in Crysis/Config folder first.

http://forum.notebookreview.com/showthread.php?t=173512&page=1143

Edit! Screens in this thread too

 

10/6/2008 2:43:51 AM - Vista 64
Beginning Run #1 on Map-island, Demo-benchmark_gpu
DX10 1280x1024, AA=No AA, Vsync=Disabled, 32 bit test, FullScreen
Demo Loops=3, Time Of Day= 9
Global Game Quality: High
==============================================================
TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 74.99s, Average FPS: 26.67
Min FPS: 20.21 at frame 1952, Max FPS: 31.79 at frame 995
Average Tri/Sec: 26256072, Tri/Frame: 984480
Recorded/Played Tris ratio: 0.93
!TimeDemo Run 1 Finished.
Play Time: 71.08s, Average FPS: 28.14
Min FPS: 19.61 at frame 1061, Max FPS: 32.91 at frame 981
Average Tri/Sec: 28014810, Tri/Frame: 995614
Recorded/Played Tris ratio: 0.92
!TimeDemo Run 2 Finished.
Play Time: 71.20s, Average FPS: 28.09
Min FPS: 19.61 at frame 1061, Max FPS: 32.91 at frame 981
Average Tri/Sec: 27945874, Tri/Frame: 994914
Recorded/Played Tris ratio: 0.92
TimeDemo Play Ended, (3 Runs Performed)
==============================================================

Completed All Tests

<><><><><><><><><><><><><>>--SUMMARY--<<><><><><><><><><><><><><>

10/6/2008 2:43:51 AM - Vista 64

Run #1- DX10 1280x1024 AA=No AA, 32 bit test, Quality: High ~~ Overall Average FPS: 28.115


Or, to save us all the trouble..and as Aeris requested a DIRECT comparison, of the non-OCd 9800m GTS vs 8800m GTX for informative purposes..here is my half of the puzzle. Same as before, settings in game to high, and res to 1280x1024. Then ran bench.. Anyhow, maybe now that you've seen mine..you'll show me yours.

 

Can anyone put everyones Crysis results side by side for easy comparison please?

 

Very nice screen shots.. DX9 has some nice hacks.

 

Looking good, actually those extra 32 SP´s of the 8800m GTX doesn´t destroy any 8800m GTS nor any 9800m GTS, it is a tad faster though. But anyway the 8800m GTS and 9800m GTS is very nice GPU´s indeed. They are basically an desktop 9600 GT

 

Agreed on the 8800m GTS VS 9800m GTS.

Nice config, those screenshots look beautiful and fantastic overall.

Thanks for the data, E-Wrecked, I will keep your post's link at handy for when someone is willing to bench a Stock 8800m GTX, or for when I get my laptop back from repairs and if I have any spare money to buy Crysis (going through a harsh situation right now, moved, old house has not been sold yet, renting a house, etc...) and do some benchmarking.

 

Hmm E-Wrecked can you overclock your 9800m GTS to the highest you can and run the Crysis benchmark again at 1920x1200. I am interested to see if a high overclock on the Core/Shaders/Mem can compensate for the 32 SP´s that isn´t there compared to the 8800m GTX.

 

just download the demo. it should work with it as well... (speculation of course)

 

47 frames huh....just ran my gtx280 and it got 35. so ummmmmmm, yo sager here i come!! can i get 45 frames from the frame gods!...lmao

edit: and my 3dmark score is like 17k

 

Hehe, I just hope it won't take days to download on my measly 1.5 Mbps / 280 Kbps Cable Modem connection. xP

Edit:

Hehe, nice framerate on the GTX 280, John, maybe 47 FPS were a bit... off?, it sounds unbelievable, maybe as Magnus said, he tweaked Crysis around to increase his FPS, who knows?

 

im sure it's the internal settings. just as i made the same mistake..then thinking i was breaking the the ultimate frames barriers only to find that my config file was that of the natural mod running at 1024x768. hahahahaha,

 

Hehe, such things tend to happen often, since such small, little settings are not really tinkered with daily!

 

johnksss well I asked him about it on the Sager forum, he hasn´t responded yet. Since first he ran with the front end that has a bug as we have concluded. Then he claimed he got the same framerate running with the .bat file. He even posted the bat file. But as I said no way he gets that at 1920x1200 DX10 High, since that is basically two GTX 260 in SLI, or not even two 8800GTX desktop in SLI gets 47-50 fps average at 1920x1200. As far as I know an 8800GTX desktop kicks the crap out of a 9800m GT

 

This is so funny!

Now your stock everything is suppose to best the overclocked everything of johnksss

O boy


johnksss scores overclocked

1280x1024

TimeDemo Play Started , (Total Frames: 2000, Recorded Time: 111.86s)
!TimeDemo Run 0 Finished.
Play Time: 73.61s, Average FPS: 27.17
Min FPS: 15.76 at frame 152, Max FPS: 31.56 at frame 350
Average Tri/Sec: 27465760, Tri/Frame: 1010841
Recorded/Played Tris ratio: 0.91
!TimeDemo Run 1 Finished.
Play Time: 70.00s, Average FPS: 28.57
Min FPS: 15.76 at frame 152, Max FPS: 31.90 at frame 342
Average Tri/Sec: 29193070, Tri/Frame: 1021689
Recorded/Played Tris ratio: 0.90
!TimeDemo Run 2 Finished.
Play Time: 70.00s, Average FPS: 28.57
Min FPS: 15.76 at frame 152, Max FPS: 31.90 at frame 342
Average Tri/Sec: 29192104, Tri/Frame: 1021736
Recorded/Played Tris ratio: 0.90
!TimeDemo Run 3 Finished.
Play Time: 70.01s, Average FPS: 28.57
Min FPS: 15.76 at frame 152, Max FPS: 31.90 at frame 342
Average Tri/Sec: 29195576, Tri/Frame: 1022030
Recorded/Played Tris ratio: 0.90
TimeDemo Play Ended, (4 Runs Performed)


both are over clocked scores.

 

Eh.. not sure which of his runs those came from.. but yeah, my stock is what it is.. if ya go back to pg. 22 you'll see that OC'd it only gained me about 4 FPS.

 

yeah, same here...my over clock and stock where like 2 to 3 frames different. i didn't quite understand that. same with going from different speeds as well...