xfx 8800gts320mbs to evga 8800gtx... is this normal ?

Stardusted

Limp Gawd
Joined
May 10, 2007
Messages
389
After my upgrade I noticed two things (all other parts remain the same). First, in colin dirt I saw little, if any, performance gains.

Then, I fired up 3dmark06, and noticed a score drop from 14300 to 12000. Even though my 8800gts was overclocked to 650/920, it shouldnt be faster than my new gtx now, should it ?

Plus, in dirt after some minutes in game it froze, then went fast ahead like hitting fast forward, then froze again, etc etc. I alt tabbed and saw temperatures at normal numbers.

Normal or RMA the damn thing ?
 
Tried the latest WHQL drivers? Also how the hell did you get 14300 with a single gts320? o_O
 
I am running the latest drivers, as to how I got 14300 well I don't really know, only thing I do know is the number.

So, say I accidentaly altered some settings and got that high number before, not that I remember doing so, but lets assume that.

Is 12000 a normal score for my system ?
 
For a stock clocked GTX that sounds right (they usually score 11000 @ stock). Also you're not gonna notice much of an improvement in dirt because it's not really making use out of the extra memory in the GTX, only the higher memory bandwidth and stream processors. Over all I'd say not to expect a very noticeable difference unless you run the game at very high resolutions. Tried to compare in any other games?
 
I'll install FEAR and run that benchmark as well. I know that it might sound stupid to worry over a benchmark number, but I'm only worried of it as a way to indicate that everything works fine. After all it is an expensive card :S

Thank you for the answers, I will post some FEAR results in a few, when I install the thing.
Again thanks, and sorry for probably being over-worried.
 
I installed FEAR and bench'ed it, it goes like :

55 min
113 avg
300 max

all settings at max, resolution 1280x1024
 
My 640MB GTS is flashed to 648/1000/1512 shader and with E6420 @ 3.625Ghz I get 11,600 in 3dmark. 14000 would only be possible with quad core and about the same OC on the GPU. In fear I get min 54, avg 109, max not sure close to 300.

A stock GTX gets about 11,500 in 3dmark with dual core - so your results with GTX sound about right.
 
I have higher clocks on both the cpu and gpu and I get around 12k+. How are you getting 14k?
 
Did you do the whole "safemode, driver cleaner, new drivers" and all when switching cards? If not, do so now. Also, if you're running Vista have you installed the hotfixes for 8800's? I'd imagine so, just checking.

3dmark score looks fine, but the FEAR benchmark looks low. I haven't run it recently though, or ever at that res, so I'm not so sure. Do you have a 1280x1024 panel? If so the GTX is something of a waste..
 
I have done all the safe mode - cleaner routine, I am running windows xp 32bit, so no vista fixes.

I know gtx is overkill for my 19'' tft, but I had the option to buy it so I took it, just to be more future proof, if anyone can be future proof in pc hardware (not the easiest thing).

As for the friend with e6600 who asks about the 3dmark06 score, quad cores as far as I know score more points than dual cores, so that explains more or less the score difference.

Anyone could post some similar system (q6600 and 8800gtx at 1280x1024) 3dmark06 results and if possible FEAR results ? If anyone could get into the trouble to do so, I would be grateful.

I already think of returning the card and getting another one just to be safe about it. If I do that I will have the choice of getting either an EVGA 8800gtx again, or a TwinTech, or an Albatron. I assume I should go after the evga again... right ?
 
Im using a Q6600. The E6600 is in my workstation. Both comps are in my sig. Q6600 @ 3400 and 8800GTS 320 @ 692/2108.
 
If that's the case then, I don't know. Thing is, shouldnt I have scored higher both in 3d and fear with my q6600 @ 3.3 and the gtx instead of a gts like you ?

Only thing I have to remind me the scores I was getting with 3dmark is a text file I had some stored in just as a point of reference. So maybe I had altered something or whatever.

But my question is still the same, is it normal ? could the 100mhz more of your q6600 make it up for the difference between the gtx and the gts in 3dmark ? Or should I just uninstall the damn thing :p
 
Back
Top