Framerate comparison
EldredP
benchmark on 3dMark2001. I also ran several games with graphics full, and
1024x768 resolution.
Computer 1(Athlon 1.2G, 384mb ram, win98se)
3dmark - 4271
GTR fps(sitting in pits) - 29
F1C(at pit exit)
San Marino - 31
Spa - 42
N2K3 TPTCC(in pits)
WG (boot) - 15
Spa - 18
Nurburgring - 22
Computer 2( AMD 1800XP, 512mb ram, win98se)
3dmark - 5463
GTR - 45
F1C
San Marino - 38
Spa - 55
Didn't get a chance to test N2k3 yet on computer 2, but I would expect a
similar increase in fps. Question is, are the 3dmark benchmark and game
framerates consistent with a computer of the second computer's configuration?
Any feedback would be appreciated.
Eldred
--
http://www.umich.edu/~epickett
Screamers League
IICC League
GPLRank -6.0 MoGPL rank +267.80
Ch.Rank +52.58 MoC +741.71
Hist. +82.34 MoH:na
N2k3 rank:in progress
Slayer Spektera lvl 72 assassin
Slayer Spectral_K lvl 38 Necro
US East
Leo Landman
> Computer 2( AMD 1800XP, 512mb ram, win98se)
> 3dmark - 5463
My 2000XP with the same card and OS gives me around 10000 3Dmarks 2001
Looks like your GF4 is sub par in both systems.
Bye,
Leo
EldredP
I wonder why mine would be so much lower...bad card? Well, it's out of
warranty now, dammit. Probably defective from the start, but I just didn't
realize it.
Argh...the hits just KEEP on coming... :-(
John Smith
"EldredP" <eld...@aol.comSPAM-OFF> wrote in message
news:20040507201236...@mb-m24.aol.com...
> >> Computer 2( AMD 1800XP, 512mb ram, win98se)
> >> 3dmark - 5463
> >
> >My 2000XP with the same card and OS gives me around 10000 3Dmarks 2001
> >Looks like your GF4 is sub par in both systems.
> >
>
> I wonder why mine would be so much lower...bad card? Well, it's out of
> warranty now, dammit. Probably defective from the start, but I just
didn't
> realize it.
> Argh...the hits just KEEP on coming... :-(
>
> Eldred
I had simular problems when I installed my gf4 4200. Try det. destroyer to
get rid of any old drivers. Then re-install new drivers, also try and reload
direct x. I get over 10000 3dmarks 2001 with an athalon 2100xp. You need a
good power supply too, 350w or better..
EldredP
>I had simular problems when I installed my gf4 4200. Try det. destroyer to
>get rid of any old drivers. Then re-install new drivers, also try and reload
>direct x. I get over 10000 3dmarks 2001 with an athalon 2100xp. You need a
>good power supply too, 350w or better..
It isn't old drivers - I did a fresh install a few weeks back on a new drive in
the same machine. I got the same values for N2k3 and 3Dmark.
But...<checks inside computer>...this power supply is only 300w...<checks
computer 2>...as is the other one. WTF - could that be slowing down the
computers? That doesn't make any sense...
I think I have another p/s around here somewhere. I got one to replace for
someone and he ended up not needing it. I just have to FIND the damn thing...
BRH
meaningless. However, I do know that 3DMark2001 yields much higher scores than the
newer version (3DMark2003?). Are you and the others running the same benchmark
program? Before you jump to conclusions, make sure that you're comparing apples to
apples.......
Bert
Peter Ives
<eld...@aol.comSPAM-OFF> writes
>>> Computer 2( AMD 1800XP, 512mb ram, win98se)
>>> 3dmark - 5463
>>
>>My 2000XP with the same card and OS gives me around 10000 3Dmarks
>2001
>>Looks like your GF4 is sub par in both systems.
>>
>
>I wonder why mine would be so much lower...bad card? Well, it's out of
>warranty now, dammit. Probably defective from the start, but I just didn't
>realize it.
>Argh...the hits just KEEP on coming... :-(
>
What AGP setting is it at. x2, x4, x8? Perhaps that's where your
slowdown is occurring.
--
Peter Ives (AKA Pete Ivington)
Remove ALL_STRESS before replying via email
If you know what's good for you, don't listen to me :)
GPLRank Joystick -50.63 Wheel -25.01
straiton
--
Steve H
"BRH @comcast.net>" <"bhoenig<no-spam> wrote in message
news:409C344F...@comcast.net...
EldredP
>>warranty now, dammit. Probably defective from the start, but I just didn't
>>realize it.
>>Argh...the hits just KEEP on coming... :-(
>>
>
>What AGP setting is it at. x2, x4, x8? Perhaps that's where your
>slowdown is occurring.
Dunno - where's that setting?
EldredP
Yes.
EldredP
>meaningless. However, I do know that 3DMark2001 yields much higher scores
>than the
>newer version (3DMark2003?). Are you and the others running the same
>benchmark
>program? Before you jump to conclusions, make sure that you're comparing
>apples to
>apples.......
Yep, we're both running 3Dmark 2001.
Peter Ives
<eld...@aol.comSPAM-OFF> writes
>>>I wonder why mine would be so much lower...bad card? Well, it's out of
>>>warranty now, dammit. Probably defective from the start, but I just didn't
>>>realize it.
>>>Argh...the hits just KEEP on coming... :-(
>>>
>>
>>What AGP setting is it at. x2, x4, x8? Perhaps that's where your
>>slowdown is occurring.
>
>Dunno - where's that setting?
>
You need to get hold of some software that will display that info.
Powerstrip is one such generic progam that works with most types of
cards. For ATI cards Rage 3D Tweak will work, but to find the info is
more complicated. You need to right-click on ATI Icon and select
settings>>ATI display settings>>then select the options tab and then
click on details. You should find the info there.
I'm sure there's other software that will also provide this info,
though.
Andi Cole
Have you updated your motherboard drivers? Latest Intel .inf files, the VIA
4in1 or whatever SIS use as necessary.
Andi.
--
Remove only one zero to reply.
"EldredP" <eld...@aol.comSPAM-OFF> wrote in message
news:20040507120804...@mb-m29.aol.com...
Biz
"EldredP" <eld...@aol.comSPAM-OFF> wrote in message
> >What AGP setting is it at. x2, x4, x8? Perhaps that's where your
> >slowdown is occurring.
>
> Dunno - where's that setting?
>
> Eldred
> --
> http://www.umich.edu/~epickett
> Screamers League
> IICC League
> GPLRank -6.0 MoGPL rank +267.80
> Ch.Rank +52.58 MoC +741.71
> Hist. +82.34 MoH:na
> N2k3 rank:in progress
> Slayer Spektera lvl 72 assassin
> Slayer Spectral_K lvl 38 Necro
> US East
I thought you were a computer guy at UofM, or somewhere around there?
Leo Landman
> But...<checks inside computer>...this power supply is only 300w...<checks
> computer 2>...as is the other one.
As is mine, and my GF4 runs fine. Other causes are much more likely to start
with.
Right click on your desktop, click on properties, click on the tab of your
video card. Somewhere in there you'll find the AGP settings. Check 4xAGP.
If you have an older Asus mobo, you may need the AMD AGP-driver, not the VIA
one.
Alternatively, go to a LAN meeting and have the boys sort it out <g>
Bye,
Leo
EldredP
>
>Have you updated your motherboard drivers? Latest Intel .inf files, the VIA
>4in1 or whatever SIS use as necessary.
>
Yes.
EldredP
I am, but you don't need to futz with AGP settings to run e-mail and Microsoft
Office...
EldredP
>Powerstrip is one such generic progam that works with most types of
>cards. For ATI cards Rage 3D Tweak will work, but to find the info is
>more complicated. You need to right-click on ATI Icon and select
>settings>>ATI display settings>>then select the options tab and then
>click on details. You should find the info there.
>
>I'm sure there's other software that will also provide this info,
>though.
I just d/l'd Powerstrip. According to it, the AGP setting is at 2x.
EldredP
>video card. Somewhere in there you'll find the AGP settings. Check 4xAGP.
I looked through all the NVidia settings, and it wasn't there. I found it by
running Powerstrip. It was at 2x, so I'll try 4x.
>
>If you have an older Asus mobo, you may need the AMD AGP-driver, not the VIA
>one.
>
>Alternatively, go to a LAN meeting and have the boys sort it out <g>
>
I'll see what I can do... :-)
EldredP
>running Powerstrip. It was at 2x, so I'll try 4x.
4x didn't help(4632 3Dmark). Oh well, thanks for trying.
Biz
"EldredP" <eld...@aol.comSPAM-OFF> wrote in message
> >I looked through all the NVidia settings, and it wasn't there. I found
it by
> >running Powerstrip. It was at 2x, so I'll try 4x.
>
> 4x didn't help(4632 3Dmark). Oh well, thanks for trying.
4X AGP setting shoudln't really matter much at all if you believe what the
tuners sitghs say about AGP speed. I guess they are correct.
Goy Larsen
EldredP wrote:
>
> >I looked through all the NVidia settings, and it wasn't there. I found it by
> >running Powerstrip. It was at 2x, so I'll try 4x.
>
> 4x didn't help(4632 3Dmark). Oh well, thanks for trying.
I've been following your "adventures" with some interest Eldred, and I'd
have to say that it does indeed sound like the vid card itself is the
culprit since you're seeing similar performance with it in 2 different
systems, but that would be the first time in my 8 years as a custom PC
builder that I come across this, and although there's always a first,
I'm still not quite prepared to write everything else off
Is there any chance of testing this video card in a system that is not
built and maintained by you ?
Yeah, I know I'm implying that you don't know your way around a computer
:-), but we all have our own way of doing things that we believe is the
correct way of doing things, and while I have a fair bit of experience
in building computers, I'm sure that some people in here would go "What
does he do that for ???" if they checked out one of my rigs, and would
prolly be right too
Why don't I buy the defective vid car theory ?
Well, a Ti4600 should easily score more than 4300(ish) in a 1.2 Athlon,
and while there's a remote possibility that the card is defective and
simply not performing up to par, why would it then score better on the
XP1800 ?
The performance increase from the 1.2 to the XP1800 is about what you'd
expect % wise, and it seems very strange that a card that is performing
below par on a 1.2 would score better on a faster CPU on a test that is
more demanding on the vid card than on the CPU, but then again computers
often defy logic....that's why it would be interesting to see what this
vid card would do elsewhere....or maybe the other way around, would it
be possible for you to borrow another vid card to test ?
Speaking of other vid cards, what vid card did you have in your studio
PC, is it powerful enough to run some tests ?
Beers and cheers
(uncle) Goy
"goyl at nettx dot no"
"A man is only as old as the woman he feels........"
--Groucho Marx--
EldredP
>I've been following your "adventures" with some interest Eldred, and I'd
>have to say that it does indeed sound like the vid card itself is the
>culprit since you're seeing similar performance with it in 2 different
>systems, but that would be the first time in my 8 years as a custom PC
>builder that I come across this, and although there's always a first,
>I'm still not quite prepared to write everything else off
>
>Is there any chance of testing this video card in a system that is not
>built and maintained by you ?
I have a Dell Optiplex GX150 borrowed from work that I could test it on.
>Yeah, I know I'm implying that you don't know your way around a computer
>:-), but we all have our own way of doing things that we believe is the
>correct way of doing things, and while I have a fair bit of experience
>in building computers, I'm sure that some people in here would go "What
>does he do that for ???" if they checked out one of my rigs, and would
>prolly be right too
>
No offense taken. Given the problems I've had, it's actually *entirely*
possible I've screwed something up. Although a standard install shouldn't
result in a system that's THIS far off where it should be... Seems like short
of defective hardware, you'd have to make an EFFORT to get the system to run
this slowly.
>Why don't I buy the defective vid car theory ?
>
>Well, a Ti4600 should easily score more than 4300(ish) in a 1.2 Athlon,
>and while there's a remote possibility that the card is defective and
>simply not performing up to par, why would it then score better on the
>XP1800 ?
More overhead with a faster CPU?
>The performance increase from the 1.2 to the XP1800 is about what you'd
>expect % wise, and it seems very strange that a card that is performing
>below par on a 1.2 would score better on a faster CPU on a test that is
>more demanding on the vid card than on the CPU, but then again computers
>often defy logic....that's why it would be interesting to see what this
>vid card would do elsewhere....or maybe the other way around, would it
>be possible for you to borrow another vid card to test ?
>
>Speaking of other vid cards, what vid card did you have in your studio
>PC, is it powerful enough to run some tests ?
The other card is a GF4MX440. I'm running 3dmark on it now. The software
didn't work on it before, but I discovered that the system didn't have D3D
acceleration enabled. I found that out when I tried to get the Ti4600 to work.
Neither 3dmark nor GTR would work before that.
Ok, the test just finished. 3704 benchmark with the MX440 and Athlon 1800XP.
Does that sound about right?
I'll also try the MX440 in the 1.2 system, as well as my old V5.
EldredP
>Computer 2( AMD 1800XP, 512mb ram, win98se)
>
>3dmark - 5463
>
>GTR - 45
>
>F1C
>San Marino - 38
>Spa - 55
I just tried my MX440 in the 1800XP system.
3dmark 3704
GTR wouldn't run. Crashed to desktop before it got to the track.
F1C:
San Marino - 34
Spa - 38
Not sure why there was such a big difference between the tracks with the
Ti4600, but not with the MX440...
Goy Larsen
EldredP wrote:
>
> I have a Dell Optiplex GX150 borrowed from work that I could test it on.
I'd give it a go there too
> No offense taken. Given the problems I've had, it's actually *entirely*
> possible I've screwed something up. Although a standard install shouldn't
> result in a system that's THIS far off where it should be... Seems like short
> of defective hardware, you'd have to make an EFFORT to get the system to run
> this slowly.
Some viruses run constantly in the background eating up CPU resources, I
had one customer complaining about her system being slow as a turtle, it
had 7 different viruses and one of them was running constantly in the
background running her CPU at 100% no matter what
> More overhead with a faster CPU?
True.....but if the vid card is the limiting factor then more CPU
overhead shouldn't give you a 25% increase, I know for instance that my
gaming rig is vid card limited, increasing the CPU speed from the stock
3.2 to 3.8 only gives me a few hundred points in 3Mark01, oh well, just
ponderings on my part
> The other card is a GF4MX440. I'm running 3dmark on it now. The software
> didn't work on it before, but I discovered that the system didn't have D3D
> acceleration enabled. I found that out when I tried to get the Ti4600 to work.
> Neither 3dmark nor GTR would work before that.
> Ok, the test just finished. 3704 benchmark with the MX440 and Athlon 1800XP.
> Does that sound about right?
Yeah....that sounds fairly close, I'd say that should score in the
3500-4500 range and would be fairly similar on both systems as it will
be the limiting factor
> I'll also try the MX440 in the 1.2 system, as well as my old V5.
It will be interesting to see...
EldredP
>
>EldredP wrote:
>>
>> I have a Dell Optiplex GX150 borrowed from work that I could test it on.
>
>I'd give it a go there too
>
Ok, that didn't work. The Dell is a 'slimline' case. I figured I could just
leave the cover off while the card was in it, but the backplane gets in the way
of plugging in the monitor. Crap...
>> No offense taken. Given the problems I've had, it's actually *entirely*
>> possible I've screwed something up. Although a standard install shouldn't
>> result in a system that's THIS far off where it should be... Seems like
>short
>> of defective hardware, you'd have to make an EFFORT to get the system to
>run
>> this slowly.
>
>Some viruses run constantly in the background eating up CPU resources, I
>had one customer complaining about her system being slow as a turtle, it
>had 7 different viruses and one of them was running constantly in the
>background running her CPU at 100% no matter what
Virus checked, ad-aware checked, spybot checked. A brand new install on a
different HD yielded the same numbers. I'm *very* confident it isn't a virus
causing the slowness.
>
>> More overhead with a faster CPU?
>
>True.....but if the vid card is the limiting factor then more CPU
>overhead shouldn't give you a 25% increase, I know for instance that my
>gaming rig is vid card limited, increasing the CPU speed from the stock
>3.2 to 3.8 only gives me a few hundred points in 3Mark01, oh well, just
>ponderings on my part
I tweaked the BIOS a bit, and got a 5300 or so benchmark. I set it to the 'top
performance' setting. I'll still try the other cards in my old system.
Eldred
Peter Ives
<an...@eloc00.co.uk> writes
>Eldred,
>
>Have you updated your motherboard drivers? Latest Intel .inf files, the VIA
>4in1 or whatever SIS use as necessary.
>
Actually, what type of chipset does your motherboard use, Eldred? I've
just read that VIA chipsets can severly slow down any 3dmark scores
because their northbridge memory controllers are only single DDR,
whereas SIS and Intel offer dual controllers.
EldredP
>just read that VIA chipsets can severly slow down any 3dmark scores
>because their northbridge memory controllers are only single DDR,
>whereas SIS and Intel offer dual controllers.
VIA, recent drivers. I think someone else had the same board though, and had a
better benchmark number.
Swervo
>> Actually, what type of chipset does your motherboard use, Eldred?
>> I've just read that VIA chipsets can severly slow down any 3dmark
>> scores because their northbridge memory controllers are only single
>> DDR, whereas SIS and Intel offer dual controllers.
>
>
> VIA, recent drivers. I think someone else had the same board though,
> and had a better benchmark number.
>
> Eldred
Lurker, posting for the first time here.
I had a Athlon 1800XP with a VIA KT-133A based motherboard that was
getting horrid framerates in just about everything. You said earlier
that you had updated the VIA 4in1 drivers, I actually found it nearly
impossible to install the AGP drivers, they'd make nearly every game
using D3D crash. Until the game did crash though (usually took a
couple of minutes of gameplay), I was seeing a 3x speed increase.
Running OpenGL in NR2003 helped, but in GPL I got complete video
corruption, so that wasn't an option either.
After just endless annoyance, I finally ditched the mobo and got a KT333
based one instead, despite my misgivings over VIA-based chipsets. I
went from about ~5,500 3DMark2001 scores to over 10,000. 4in1
installed fine, and things ran nicely. GTR2002 and the like still
struggled at high detail settings, but they ran great with some of the
effects turned down.
I got the distinct impression from some of those that I talked to that
some of the KT-133A chipsets were a little dodgy, and that I happened to
get ahold of one of those. I honestly can't see a motherboard upgrade
without changing the CPU or the video card making the difference it did
without a screwy motherboard in the first place.
Since then, I stuck a Radeon 9800 Pro in there, which has been nice, but
I still got very acceptable framerates with the GF4 Ti4200 that I had in
there.
I realize this is a horrid suggestion when you want to avoid spending
money, but a decent mobo isn't supremely expensive. While I'm no
expert at building systems, I can hold my own when it comes to my
hardware, and the KT-133A setup was nearly the death of me. I haven't
regretted getting away from it since I did it.
'Course, if you don't have a KT-133A chipset, then completely ignore the
post you just read ;)
EldredP
Thanks for 'de-lurking'...<g> I know it's a VIA chipset, but I'm not sure
about the KT-133A part. Hang on...
I just did a google search on what I *think* the mobo number is(GA-7zx), and
found a German site. I can't speak German, but I'm pretty sure that one of the
people says it has a KT-133 chipset. So I may have the same problem that you
encountered. I guess I'll have to see if I can free up some money for at least
a new mobo. Time to find a second job...
Swervo
> Thanks for 'de-lurking'...<g> I know it's a VIA chipset, but I'm not
> sure about the KT-133A part. Hang on... I just did a google search on
> what I *think* the mobo number is(GA-7zx), and found a German site.
> I can't speak German, but I'm pretty sure that one of the people says
> it has a KT-133 chipset. So I may have the same problem that you
> encountered. I guess I'll have to see if I can free up some money
> for at least a new mobo. Time to find a second job...
No problem :) I was so incredibly frustrated by my problems with the
board, so I hope I helped. It was almost enough to get me to swear off
VIA, but then it was so much cheaper at the time to go VIA vs. nForce
that I ended up going VIA again, and haven't had any problems.
I couldn't find KT3 Ultra2's (the replacement board I got) on sale, but
this looks like their new sorta standard mobo:
http://www.newegg.com/app/ViewProductDesc.asp?description=13-130-445&depa=0
Not *too* unreasonably priced...
Swervo
And, in my own bad form, responding to my old post...
Just dropped an Athlon 2700+ XP in the same mobo tonight. I can now
watch the run down the grid at the beginning of an F1C race at a decent
framerate. The proc was about $120 (bought it from a store that's been
good to me, despite them being a good $30-$40 more than online dealers),
and while the max framerate has barely jumped on GTR (I'm up to about
62fps vs. 55fps), the framerate when lots of cars are on the screen at
once is astronomically higher.
EldredP
>And, in my own bad form, responding to my old post...
>
Not a problem.<g>
>Just dropped an Athlon 2700+ XP in the same mobo tonight. I can now
>watch the run down the grid at the beginning of an F1C race at a decent
>framerate. The proc was about $120 (bought it from a store that's been
>good to me, despite them being a good $30-$40 more than online dealers),
>and while the max framerate has barely jumped on GTR (I'm up to about
>62fps vs. 55fps), the framerate when lots of cars are on the screen at
>once is astronomically higher.
>
Ok, thanks for that info. I'm still looking at options...