I'm already switched over to the new drivers, not going to bother going back to test resolutions I don't game at;)Quote:
Originally Posted by Arkangyl
Printable View
I'm already switched over to the new drivers, not going to bother going back to test resolutions I don't game at;)Quote:
Originally Posted by Arkangyl
Well I`m sure you got pretty nice setup there.
But we need those tests to proove the actual increase os dual-core with this driver.
So pretty please make some test for fellow guys here.
Thank you.
I tryed these quick with BF2, and they suck big time. I went from 85-100fps to 25-35fps and the shadows are just wacked. They jump all over as you walk by something casting one. I didn`t bother to run any benchs, I saw the horrible performance in the only gfame i`m playing right now, so I rolled back to the 78.01s, and yeah I have a dual core and 7800s in SLI.
I'm not having any of those issues in bf2. Did you use DriverCleaner to dump all traces of the old drivers before installing the new ones? BTW, what are you using to determine fps in bf2?Quote:
Originally Posted by Rukee
If this was directed at me I can oblige you.:)Quote:
Originally Posted by Cooper
I'll run various AA/AF configs at 1024 and 1280 in Doom3. Can anybody tell me how to run a BF2 bench?
If you could make those tests it`d be great.
But we need a comparison between single-core and dual-core with equivalent clocks with this new driver.
The drivers have been confirmed to be buggy. Lots of stuff aren't rendered, which is where the huge performance increase comes from.
Confirmed by who? In the thread over at Futuremark, a user with a 4800+ setup claimed that there was no image quality loss between this new set, and the previous one he had in Doom III; but there was a performance boost.Quote:
Originally Posted by Shadowmage
There's been similar reports over at Anandtech forums aswell.
<<< Dual Core and 7800GTX.. if someone can PM me how they would want the benches to be run.. i really only bench with AQ3,3dMark and SPi so im not familiar with benching through games.
however i have the following games.
FarCry,Bf2,DoomIII,CS:S,HL2 so if someone would like me to contribute with a GTX i'd be more than happy i just need to know how
Yes I did, and I use fraps for fps in game.Quote:
Originally Posted by ErikaeanLogic
Okay, so here is a comparison between the WHQL 78.01 and beta 81.26 NVidia drivers. The following were constant between all tests:
Athlon64 X2 3800+ @ 2700MHz
2x1GB Crucial Ballistix PC4000 @ 270fsb, 3-3-3-7
BFG 7800GT OC @ 485/1200
74GB WD Raptor (system and apps)
Doom3 for all demos using the "timedemo demo1 1" command
8x Anisotropic Filtering
High Quality setting in Doom3
Ready?
Reso/AA/78.01/81.26
1024/no/101.2/106.6
1024/2x/100.8/105.8
1024/4x/94.5/98.2
1280/no/99.6/104.4
1280/2x/92.9/96.0
1280/4x/75.2/76.9
1600/no/91.2/94.4
1600/2x/78.4/80.0
1600/4x/57.0/57.5
1920/no/84.5/86.7
1920/2x/69.6/70.6
1920/4x/48.9/49.3
So, I have to say that even I was a bit surprised. The fact that there is a fairly uniform 2-5% increase across the resolutions seems counterintuitive to me. Also, I must have uninstalled and reinstalled the 78.01 and 81.26 drivers 3 times the day of the 81.26 release and got a 4fps difference at 1920/noAA and only got a 2fps diff tonight, so that's a bit confusing, too. I don't see the 20-40% increase in Doom3 fps, which didn't surprise me. I don't notice much of a difference in video quality between the two driver sets so I can't figure out what all the fuss is about there. Anyway, I hope that this helps:)
Any driver that "Exploits Dual Core" is simply using the processor to do the work the card(s) should be doing, one of the reasons I'm not a fan of Nvidia. Just look at some '01 results and you will see how inefficient an Nvidia card is compared to ATI, they use alot more system power while doing the same things.
How about some single core vs dual core benches using different drivers? That's the only way to prove that there are actual dual core tweaks, which I'm still skeptical of.
As ErikaeanLogic's benches show, the low res gains are the same as the high res gains. If it really had dual core tweaks, low res gains should be higher since the CPU is the bottleneck in that situation.
this is really interesting to me.. can someone direct me in a way to figure this out.. as stated above im DualCored and GTX'd...
any games that you'd like to see it tried on???? also should i use FRAPS? or an actual benchmark.. i can test with AQ3.. but there is no way for me to validate since the ARC is down....
I tryed them again and got horrible results, BF2 has blue mountains, and the shadows are just wacked. They jump around for some things, like this tank, but notice the jeep doesn`t have one. Also some bushes flash back and forth like between two colors or something. The 78.01 are fine and the shadows are normal.
http://img.photobucket.com/albums/v6...-25shadows.jpg
http://img.photobucket.com/albums/v6...emountains.jpg
People, just remember these are ALPHA drivers! Not even beta, but alpha (according to Nvidia PR). The final revision will be much better assuredly..
I dont know if its already posted or not but VR-Zone also has done new driver testing. Its going good with these drivers.
Yeap they didQuote:
Originally Posted by harshal
But... I haven`t found anything good there.