WHQL 52.16

[benjamat]

Tried these drivers out briefly.

They would not install properly over previous drivers ( not showing side window contents) but when system cleaned of old drivers it installed perfectly.
Added "Coolbits" reg hack for overclocking and tried 3dmark2001

Detonator 44.03 - 11403
Forcewear 52.16 - 9896 - 87% performance

Have not run game benches yet but it seems clear that either previous drivers were jigged by nvidia or with GF4 range these drivers are no good. I suspect that these drivers deliver genuine scored unlike some others so choice is yours, if you believe this to be true then install 52.16 as latest WHQL otherwise find your best detonator driver.:frown:

 

[Jeff7181]

You've gotta be missing something. My old TI4200 at stock speeds never even scored that low with a stock speed XP1700... check to make sure vsync is off, and the rest of the usual stuff like AA and AF.
Most other people are seeing increased scores even with an old GeForce4. Mind you, these Forcewear 52.16's are designed and written for the Geforce FX line.
But either way... something else must be different, because seeing the score drop that much is not normal, double check all your settings again.

 

[benjamat]

Thanks for input

Checking on 3dmark website I see average score for systems with similar processor speed is 7994 so with my score of 11403 with detonator drivers this does not seem too bad - 142%

I do notice that half of the results between 11400 and 11500 come using 4041 drivers with 4072 coming a distant second.

I did not use these drivers but the newer 45.23

Looking at XP1700 scores they do seem as good or better than the 1.6 P4 clocked at 133 on average.

Appreciate any thoughts.

 

[ticktack]

Maybe, can you guys check if this version had problems in running your fav. games? As I did encountered with this 52.16 d/l from nVidia, and before, I'm also using this 45.23 and gaming with Midnight ClubII is very excellent with my previous Detonator. Unfortunately, yesterday when I ran the game, the far distant objects such as buildings and etc are all greyed or blackened out and images only becomes clearer is when the vehicle is approaching them, thus, I find it rather weird as I thot maybe, my system has hogged up too many memory resources thus producing this effect, but then even reboot, and re-entering the game, resulted the same. If you have idea about what I am talking about, well, it's like system memory not enough to produced enough clear gd quality images that sort of examples. However, I decided to re-install my older Detonator 45.23, and astonishingly, all is perfect, the distant building are all visible with the actual intended "fog" or "misty" types of image effects, thus, I really, have no idea why the latest has issue concerning this game and maybe, all someone has this game can try it out and share your experiences with this 2 drivers. Now, I am sticked with using 45.23 as I still wish to play this game rather than using 52.16 whereby, running 3d mark2003 doesn't has any differences with 45.23. My system spec is listed as below and hopes someone can shed "clearer" images into this issue. Thank you.

 

[zogg]

I tried those drivers 52.16 and could only play Unreal tounament for about 2 minutes then the system rebooted by itself. Its a good thing that xp has that rollback drivers feature.

 

[Insomniak]

Originally posted by: benjamat
Tried these drivers out briefly.

They would not install properly over previous drivers ( not showing side window contents) but when system cleaned of ikd drivers it installed perfectly.
Added "Coolbits" reg hack for overclocking and tried 3dmark2001

Detonator 44.03 - 11403
Forcewear 52.16 - 9896 - 87% performance

Have not run game benches yet but it seems clear that either previous drivers were jigged by nvidia or with GF4 range these drivers are no good. I suspect that these drivers deliver genuine scored unlike some others so choice is yours, if you believe this to be true then install 52.16 as latest WHQL otyherwise find your best detonator driver.:frown:

IIRC, 44.03 had IQ issues and AF "optimization". 52.16 does not. That would probably account for the decreased scores: IQ fix.

But seriously, why do you use benchmarks? They're so useless. Test your performance in real world games.

 

[benjamat]

Tested in two games, Quake 3 and UT 2003.

Both games show about 5% reduction at 1024/768 getting closer as resolutions go higher.

I know 52.16 is designed for FX cards but read somewhere that GF4's would benefit also.