• Notice regarding image hosts and attachments. See here.

  • Competitor rules

    Please remember that any mention of competitors, hinting at competitors or offering to provide details of competitors will result in an account suspension. The full rules can be found under the 'Terms and Rules' link in the bottom right corner of your screen. Just don't mention competitors in any way, shape or form and you'll be OK.

9800GTX Benched on VR ZONE

http://www.unihw.com/html/24/n-124.html

single card pre-test,

will finish dual-sli & tri-sli soon after.


9800GTX Spec：
vGPU=1.2v~1.21v idle, 1.26~1.27v load
vRam=2.0v

ambient=18c，GPU idle=35c，load=53c（100% fan speed）

overclking results summary：
1）Driver 174.20 might have some bug，3Dmark03 result 20% lower than normal，3Dmark05/06 ok.
2）heat same as 8800gts (better cooler but 0.02v vGPU higher)
3）GPU core overclking better (better PCB/PWM/Cooling)
4）vRam overclking better (0.8ns, better PCB/PWM)
5）perf_lvl same as 8800gts clk 2 clk. (only deafault clk higher)
Click to expand...




For ref next to a GTS: 8800GTS @ 850/1150MHz E6850 @ 4275mhz

 
Last edited:
bench at 1920x1200 with 4-16x aa and I might be interested in the results, 1280x1024 is a seriously stone age res with the power of today's graphics cards.
 
WatchTower said:
I'm sure my 8800GT could be hitting 16k if I was able to clock my Q6600 to 4ghz
Click to expand...

My unmodded GT got 16.7K with a 4GHz quad, not that its compares to the GTX's dual core score. I've added a 8800GTS with similar clocks and a dual core to compare. Note its a wolfdale they're using so a bit quicker than a E6850 clock for clock.
 
Last edited:
Gerard said:
3d mark score is unimpressive to say the least. Saying that it was done on a dual core, quad would obviously had bumped that score up some.
Click to expand...

Which just goes to demonstrate what a worthless heap of ***** 3DMark06 is. The results get a boost from using a quad-core yet the actual benefit is negligable in the vast majority of games.

Combined with the fact that all "standard" 3DMark06 results are at 1280x1024 without any anti-aliasing, which is hardly representative of current gaming averages, just confirms that it's an outdated piece of junk which needs replacing pronto.
 
It's their decision what to call it - they wouldn't release it as "9800GTX" if it was in fact barely better than the other cards.
 
helmutcheese said:
Hmm, well fact is 3DMARK06 is set to run by default at 1280x1024, so accept it or dont use it to compare to others.
Click to expand...
Yes but you can change it.

I bench and game at 1920x1200 with 4x or 16x aa. Vertigo1 has hit the nail on the head, it's so badly out of date and out of touch with anything it doesn't in any way reflect a systems real world game performance.
 
straxusii said:
Oh dear 512MB, that scotches Fuds rumour it was 768MB
Click to expand...


It always been 512mb. Fud are just fools who got done.



Why is everyone saying is this a joke... this canny be right etc etc... its just an overclocked GTS, what more do you expect. It should do well in SLI if the 9600GT sli performance is anything to go by... but it will prolly have a silly price tag and SLI will be out of the question for most people.
 
1280x1024 may be the default for 3dMark06, but it is pretty much pointless with the modern cards.

You only really get to see the top end cards (and SLI / Crossfire setups) show their true worth when you get to high resolutions with AA and AS enabled.

I'm sure this'll all be rectified by the release of Vantage, though.
 
Bonjour said:
1280x1024 may be the default for 3dMark06, but it is pretty much pointless with the modern cards.

You only really get to see the top end cards (and SLI / Crossfire setups) show their true worth when you get to high resolutions with AA and AS enabled.

I'm sure this'll all be rectified by the release of Vantage, though.
Click to expand...

no, it is not pointless. Running at a default of anything above 1280x1024 would block many people from using it.

Just because you game at a high resolution doesnt mean eveyrone else does. to have a comparable benchmark it must be run at the lower end of the scale to fit in as many people as possible.

I will check the valve thingy, but i think 1280x1024 is the most commonly used res, on steam at least!

EDIT: Sauce

Primary Display Resolution (1385540 Users)
800 x 600 23,369 1.69 %
1024 x 768 441,151 31.84 %
1152 x 864 71,967 5.19 %
1280 x 960 548,246 39.57 %
1440 x 900 101,675 7.34 %
1600 x 1200 23,737 1.71 %
1680 x 1050 124,770 9.01 %
1920 x 1200 31,439 2.27 %
Other 19,186 1.38 %

So at 1280x1024 they have already kicked out over 32% of the market. losing 1280x1024 removes another 45%!
 
Last edited:
1280x1024 without AA isn't just useless, it's downright misleading.

With NVidia reducing the memory bandwidth on the latest G92 cards compared to some of the G80s, this can seriously hamper gaming at higher resolutions with anti-aliasing. A 3DMark06 score at default settings could well give a false impression of the card's performance and higher resolutions.

Actually, I'd be very interested in some 3DMark06 scores at 1920x1200 with and without 4xAA from people with 512Mb GT and GTS cards to see how much they really do suffer at these resolutions. If anyone has one of these cards and a registered copy of 3DMark06 and feels the urge to contribute some figures...

EDIT: Skeeter - I heard what you're saying and you're right - the majority of gamers do still game at lower resolutions. I still think the average res is climbing, and 1680x1050 is gaining ground quite rapidly but you are correct, the average still seems to be around the 1280x1024 mark.

This is obviously why NVidia have felt they can reduce the memory bandwidth of the latest cards and jack up the "processing power" instead. At these average resolutions the overall effect is a positive one as the processing power increase makes itself felt whilst the memory bandwidth is still sufficient.

The point I'm making is that not everyone does game at the lower resolutions. Just as we shouldn't assume there are loads of 1920x1200 gamers, we also shouldn't assume that there are none. Those of us with 24" monitors who want to game at higher resolutions with anti-aliasing enabled need to be aware that 3DMark06 benchmarks at the default settings are next to useless as the cards in question could well run into serious memory bottlenecks as the resolution is increased.

I think it's time more sites published 3DMark06 benchmarks not only with the standard settings but at higher ones with AA too.
 
Last edited:
Skeeter said:
no, it is not pointless. Running at a default of anything above 1280x1024 would block many people from using it.

Just because you game at a high resolution doesnt mean eveyrone else does. to have a comparable benchmark it must be run at the lower end of the scale to fit in as many people as possible.

I will check the valve thingy, but i think 1280x1024 is the most commonly used res, on steam at least!

EDIT: Sauce

Primary Display Resolution (1385540 Users)
800 x 600 23,369 1.69 %
1024 x 768 441,151 31.84 %
1152 x 864 71,967 5.19 %
1280 x 960 548,246 39.57 %
1440 x 900 101,675 7.34 %
1600 x 1200 23,737 1.71 %
1680 x 1050 124,770 9.01 %
1920 x 1200 31,439 2.27 %
Other 19,186 1.38 %

So at 1280x1024 they have already kicked out over 32% of the market. losing 1280x1024 removes another 45%!
Click to expand...

All I was saying was that the top-end cards / setups are best suited for gaming at higher resolutions.

You'll probably find that the vast majority people considering purchasing one of these cards actually do game at resolutions higher than 1280x1024.
 
You see, that just shows that 3DMark at 1280x1024 tells you more about the CPU than it does about the latest graphics cards.
 
You must log in or register to reply here.
Back
Top Bottom