Nvidia 56.72 Driver Performance

[Guest]

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

I'm a bit slow in upgrading to the 56.72's from the 53.03's.
I noticed a performance drop in the 3DMark2001SE tests and wondered if this
should be expected?
System: AMD Barton 2500 running at XP3000 speed, Asus A7N8X Deluxe, 1GB
Crucial DDR400

3DMark03
gf fx5900 clock475 ram950 5865 nv56.72
gf fx5900 clock475 ram950 5841 nv53.03

3DMark2001SE
gf fx5900 clock475 ram 950 56.72 14958
gf fx5900 clock475 ram 950 53.03 15502

Is it something to do with driver optimisations being removed?

Thanks,
Graham.

 

[tiger]

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

I turned back to the old 53.03 drivers, 'cause some games (Runaway) didn't
start with the new 'forceware' drivers! I don't want better
benchmark-results when I cann't play 1 year old games anymore...
System XP Prof.
AMD 2600+
MSI FX5900XT VDT128
512 MB DDR 400

Ciao
Tiger

"The Old Man" <spam@spam.com> schrieb im Newsbeitrag
news:40ae0112$0$25324$cc9e4d1f@news-text.dial.pipex.com...
> I'm a bit slow in upgrading to the 56.72's from the 53.03's.
> I noticed a performance drop in the 3DMark2001SE tests and wondered if
this
> should be expected?
> System: AMD Barton 2500 running at XP3000 speed, Asus A7N8X Deluxe, 1GB
> Crucial DDR400
>
> 3DMark03
> gf fx5900 clock475 ram950 5865 nv56.72
> gf fx5900 clock475 ram950 5841 nv53.03
>
> 3DMark2001SE
> gf fx5900 clock475 ram 950 56.72 14958
> gf fx5900 clock475 ram 950 53.03 15502
>
> Is it something to do with driver optimisations being removed?
>
> Thanks,
> Graham.
>
>

 

[Guest]

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

"tiger" <me@privacy.net> wrote in message
news:2h6o49F9q0apU1@uni-berlin.de...
> I turned back to the old 53.03 drivers, 'cause some games (Runaway) didn't
> start with the new 'forceware' drivers! I don't want better
> benchmark-results when I cann't play 1 year old games anymore...
> System XP Prof.
> AMD 2600+
> MSI FX5900XT VDT128
> 512 MB DDR 400
>
> Ciao
> Tiger

That's a good point.

 

[Guest]

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

"The Old Man" <spam@spam.com> wrote in message
news:40ae4218$0$20512$cc9e4d1f@news-text.dial.pipex.com...
>
> "tiger" <me@privacy.net> wrote in message
> news:2h6o49F9q0apU1@uni-berlin.de...
> > I turned back to the old 53.03 drivers, 'cause some games (Runaway)
didn't
> > start with the new 'forceware' drivers! I don't want better
> > benchmark-results when I cann't play 1 year old games anymore...
> > System XP Prof.
> > AMD 2600+
> > MSI FX5900XT VDT128
> > 512 MB DDR 400
> >
> > Ciao
> > Tiger
>
> That's a good point.
>
>
>
The Old Man, You should really try out some of the leaked betas. Some of
them get a pretty big increase in realtime and benchmarking both with fps
and IQ. Yes ,they tend to have bugs but so do the WHQL's. At least with the
beta releases you know there will be bugs. Where as with the WHQL they are
not suppose to have any bugs.

 

[Guest]

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

"empireAA" <nobody@NOSPAMSPAMSUX.com> wrote in message
news:2Grrc.3577$0X2.140263@twister.tampabay.rr.com...
>
> "The Old Man" <spam@spam.com> wrote in message
> news:40ae4218$0$20512$cc9e4d1f@news-text.dial.pipex.com...
> >
> > "tiger" <me@privacy.net> wrote in message
> > news:2h6o49F9q0apU1@uni-berlin.de...
> > > I turned back to the old 53.03 drivers, 'cause some games (Runaway)
> didn't
> > > start with the new 'forceware' drivers! I don't want better
> > > benchmark-results when I cann't play 1 year old games anymore...
> > > System XP Prof.
> > > AMD 2600+
> > > MSI FX5900XT VDT128
> > > 512 MB DDR 400
> > >
> > > Ciao
> > > Tiger
> >
> > That's a good point.
> >
> >
> >
> The Old Man, You should really try out some of the leaked betas. Some of
> them get a pretty big increase in realtime and benchmarking both with fps
> and IQ. Yes ,they tend to have bugs but so do the WHQL's. At least with
the
> beta releases you know there will be bugs. Where as with the WHQL they are
> not suppose to have any bugs.

Thanks, I have tried them for some time before but got fed up with new ones
appearing so often. I try to stick to the new ones these days but having
said that didn't expect such a theoretical drop in performance for games
that are 1-2 years old. Now I have a great graphics card, I'd like to
revisit the older games and replay with advanced filtering, A.A. etc to see
the difference. All part of the fun! Still benchmarking isn't everything
though.

Graham.

 

[Dano]

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

"The Old Man" <spam@spam.com> wrote in message
news:40af4c69$0$25329$cc9e4d1f@news-text.dial.pipex.com...
>
> "empireAA" <nobody@NOSPAMSPAMSUX.com> wrote in message
> news:2Grrc.3577$0X2.140263@twister.tampabay.rr.com...
> >
> > "The Old Man" <spam@spam.com> wrote in message
> > news:40ae4218$0$20512$cc9e4d1f@news-text.dial.pipex.com...
> > >
> > > "tiger" <me@privacy.net> wrote in message
> > > news:2h6o49F9q0apU1@uni-berlin.de...
> > > > I turned back to the old 53.03 drivers, 'cause some games (Runaway)
> > didn't
> > > > start with the new 'forceware' drivers! I don't want better
> > > > benchmark-results when I cann't play 1 year old games anymore...
> > > > System XP Prof.
> > > > AMD 2600+
> > > > MSI FX5900XT VDT128
> > > > 512 MB DDR 400
> > > >
> > > > Ciao
> > > > Tiger
> > >
> > > That's a good point.
> > >
> > >
> > >
> > The Old Man, You should really try out some of the leaked betas. Some of
> > them get a pretty big increase in realtime and benchmarking both with
fps
> > and IQ. Yes ,they tend to have bugs but so do the WHQL's. At least with
> the
> > beta releases you know there will be bugs. Where as with the WHQL they
are
> > not suppose to have any bugs.
>
> Thanks, I have tried them for some time before but got fed up with new
ones
> appearing so often. I try to stick to the new ones these days but having
> said that didn't expect such a theoretical drop in performance for games
> that are 1-2 years old. Now I have a great graphics card, I'd like to
> revisit the older games and replay with advanced filtering, A.A. etc to
see
> the difference. All part of the fun! Still benchmarking isn't everything
> though.
>
> Graham.
>
>

the newer drivers are said to eliminate the "NVIDIA performance boost" in
certain benchmarking..i.e.3d mark. it has been widely publicized that NVIDIA
drivers of past falsely boosted scores in madonions product, also many
graphics card makers now view 3d mark as a elaborate demo not a real gaming
benchmark(funny, swear at it when it is not in your favor, swear by it when
they are)...so the drop in performance is real world results so to speak,
not marketing results to boost sales as in the past.

 

[Guest]

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

I wouldn't sweat it. The 3dmark2001se performance drop you have
indicated doesn't seem significant in any way at all, and the numbers
are sweet anyway, my TI4200 won't break 10,000!

Sounds like everything is fine.

 

[Guest]

Archived from groups: alt.comp.periphs.videocards.nvidia (More info?)

> >
> >
>
> the newer drivers are said to eliminate the "NVIDIA performance boost" in
> certain benchmarking..i.e.3d mark. it has been widely publicized that NVIDIA
> drivers of past falsely boosted scores in madonions product, also many
> graphics card makers now view 3d mark as a elaborate demo not a real gaming
> benchmark(funny, swear at it when it is not in your favor, swear by it when
> they are)...so the drop in performance is real world results so to speak,
> not marketing results to boost sales as in the past.

bri-linear filtering can be switched of with an advanced feature in
the driver. Normally it's on but it can be switched off!