• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Driver Roundup, By: Icedragn

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

icedragn

Member
Joined
Jan 20, 2005
Location
Ontario, Canada
The following charts are based on the following system:

AMD 64 3500+
1gig OCZ DC DDR
160gb SATA HDD
Leadtek 6800LE @ Stock 8x1,4vp 300/700
Windows XP SP2

3DMark 05 = Default Settings (only game tests are ran)

CSS Stress Test:
Resolution: 1024x768
Models - High
Textures - High
Shaders - Low
Water - Simple
Shadows - High
AA - 2x
AF - Trilinear
HDR - Full (dunno if this matters for stress test??)

82.12 Tweaked = Z-Tweaked
81.98 Tweaked = Xtreme-G
81.85 Tweaked = Z-Tweaked

nvidiadrivers3dmark052lx.png


nvidiadriverscss4da.png


As you can see from the graphs above, the z-tweaked 81.85's are the fastest drivers, for my system at least. You can see in 3DMark 05 they are almost tied with the 82.12's, but really they are 3 3dmarks higher, but you really notice the difference in CSS, they pull ahead by 1.52fps. All in all, the z-tweaked drivers offer great performance and still have good enough image quality (that i can tell), and if it looks fine to me, why not get a nice boost in performance? You can get z-tweaked drivers from http://downloads.guru3d.com or http://www.thezproject.org. You can also see that the 81.98 nVidia drivers are better than the "tweaked" version from Xtreme-G LOL. In CSS the nVidia 81.98's are 0.3fps faster, and in 3dmark 05 they are 12 3dmarks faster, this is such a small margin, but still funny that a "tweaked" version actually runs slower than the originals, even if just marginally.

HAPPY GAMING! =]

**3DMark 06 Tests To Come (LOL)**
 
Last edited:
ok thats nice but how does it do with image quality. because most "high performance" drivers do away with high quality images and well people that buy high end cards bought the card for a couple of reasons, to make the game play smoother and to have a better visual effect. so lowering the image quality to gain fps and points is just waisting a high end card in my opinion
 
liftedcj7on44s said:
ok thats nice but how does it do with image quality. because most "high performance" drivers do away with high quality images and well people that buy high end cards bought the card for a couple of reasons, to make the game play smoother and to have a better visual effect. so lowering the image quality to gain fps and points is just waisting a high end card in my opinion

i don't see a difference in CSS compared to the older drivers.

edit:
also a 6800LE isn't high end =], but the 11FPS boost is very nice, and like i said, i don't notice a difference.
 
liftedcj7on44s said:
ok thats nice but how does it do with image quality. because most "high performance" drivers do away with high quality images and well people that buy high end cards bought the card for a couple of reasons, to make the game play smoother and to have a better visual effect. so lowering the image quality to gain fps and points is just waisting a high end card in my opinion


I agree. Tho right now my A64 is down as i'm waiting on a new cpu, but once it does get here I'll try these "Z-Tweaked' drivers and see for myself. Benchmarks are one thing, real world performance/results is what counts.
 
Nebulous said:
I agree. Tho right now my A64 is down as i'm waiting on a new cpu, but once it does get here I'll try these "Z-Tweaked' drivers and see for myself. Benchmarks are one thing, real world performance/results is what counts.

which is why i used CSS, forgot about the Quake 4 becnhmark but i don't feel like doing that all over again :p, maybe tomorrow.
 
icedragn said:
which is why i used CSS, forgot about the Quake 4 becnhmark but i don't feel like doing that all over again :p, maybe tomorrow.

Well once my A64 is up and running again, i'm going to run these thru the mill and test evry game benchmark i have plus run all my games and see. I'll post my findings.
 
Running more tests as we speak, this time i'm not using the XG 81.95 and the 71.89 NV drivers, and i have added Z-T 81.85/NV 81.98. Will run FEAR and Quake 4 benchmarks.
 
Try using the same Driver verison, but just different modifications of it for a more accurate portrayal. Those Z-Ts do look like they yield a huge performance gain.
 
darksparkz said:
Try using the same Driver verison, but just different modifications of it for a more accurate portrayal. Those Z-Ts do look like they yield a huge performance gain.

don't see any 82.12 from XG, and nvidia is only at 81.98. I'll see if i can find the leaked 82.12 from nv. Also i'm not going to be able to do Q4 benches, it requires the full game.. so i'm doing a Doom 3 Bench instead.

edit:
dammit, d3 demo is for the full game too.

Can i benchmark using the demos? like a built in feature? otherwise i'm only going to add FEAR.
 
Updated, ran the tweaked drivers against their nVidia counterparts.

Still only used CSS/3DMark 05. Fear demo didn't have a "Test Settings" option, so no benchmark there, Doom 3 requires the full game, and quake 4 requires the full game. Was going to use the FF11 benchmark but it takes like 5 min+ just to do it once, that's too long :p
 
Back