Detenator 44.03's

[Schneider]

UPDATED!!

These drivers where used in Anandtech's review of the 5900 Ultra and it seems
the main performance boost is given to the FX 5*** series. Bear in mind that these
drivers are only beta's. You can get them at Nvmax
They are dated at 11/5/03 so they are relatively new.

Test Bench:

P4 2.26 @ 2.4 (141 FSB)
DDR 282 2.5,7,2,2
Gigabyte 8IR533
Geforce4 MX 420
High Performance, Performance and Quality are the Intellisample options
ADDED 4x AA performance results

3D Mark 2001SE:

High Performance : 4228
Performance : 4082
Quality : 4072
Omega 43.45 perf: 4087

3D Mark 03:

High Performance : 212
Performance : 202
Performance 4xAA: 78
Quality : 201
Omega 43.45 perf: 195

Unreal Tournament 2003:

Anatulus Flyby min/avg fps:

High Performance: 35.741192 / 65.434502
Performance : 21.802521 / 65.420486
Performance 4xAA:11.546674 / 18.586821
Quality : 20.009563 / 65.429062

Asbestos Flyby min/avg fps:

High Performance: 43.228012 / 120.007416
Performance : 35.126228 / 119.990173
Performance 4xAA:17.975182 / 34.003090
Quality : 39.802372 / 119.963737

Citadel Flyby min/avg fps:

High Performance: 6.754090 / 55.284843
Performance : 14.202627 / 55.444191
Performance 4xAA:5.189257 / 18.541336
Quality : 14.176000 / 55.337090

Crytek's X-Isle min/avg fps:

High Performance : 19/33.03
Performance : 19/32.94
Quality : 19/32.54
Omega 43.45 perf: 19/ 33.33

Post ur results....as my MX420 isn't exactly the best to see performance from
Will update later with more benchy results

Does anyone know how to benchmark using Jedi Knigh 2: Jedi Outcast ver1.04 and Aquanox ver1.18??

 

[Intelman07]

They are dated at 11/5/03 so they are relatively new

Uhhh we aren't in December of 2003 yet........

 

[JZilla]

I think Schneider is european.

And I also got a GF4 MX420 and I have noticed no performance increase going from 30.82 to 41.09 that I am using now. I think the card is already delivering all it can.

 

[nemesismk2]

I'll see what kind of performance increase (if any) I get with my fx5200 when I finish work tonight and post results. (I will also re-benchmark the fx5200 when my xp 2500 (barton) arrives later in the week)

 

[nemesismk2]

Had enough time to run a few test before my lunch:-

athlon xp 1700
asrock sis746fx motherboard
xfx 128mb geforce fx5200
512mb ddr266 memory

3dmark 2001 se = 43.51=6552 44.03=6612
3dmark 03 = 43.51=1346 44.03=1375

The biggest increase with 3dmark 03 is the single texturing fillrate which increased from 465.7 MTexels/s with the 43.51 drivers to 610.4 MTexels/s with the 44.03 drivers.

The biggest loss is with 4xfsaa enabled, the 44.03 drivers result in a large performance hit compared to the 43.51 drivers. Although to be fair to Nvidia these drivers don't even list the FX5200 as being supported so I'll wait for the official drivers before doing anymore benchmarking.

 

[Jeff7181]

Originally posted by: JZilla
I think Schneider is european.

And I also got a GF4 MX420 and I have noticed no performance increase going from 30.82 to 41.09 that I am using now. I think the card is already delivering all it can.

A 1 Ghz Duron could probably push that card to it's limits.

 

[Jeff7181]

Originally posted by: nemesismk2
Had enough time to run a few test before my lunch:-

athlon xp 1700
asrock sis746fx motherboard
xfx 128mb geforce fx5200
512mb ddr266 memory

3dmark 2001 se = 43.51=6552 44.03=6612
3dmark 03 = 43.51=1346 44.03=1375

The biggest increase with 3dmark 03 is the single texturing fillrate which increased from 465.7 MTexels/s with the 43.51 drivers to 610.4 MTexels/s with the 44.03 drivers.

The biggest loss is with 4xfsaa enabled, the 44.03 drivers result in a large performance hit compared to the 43.51 drivers. Although to be fair to Nvidia these drivers don't even list the FX5200 as being supported so I'll wait for the official drivers before doing anymore benchmarking.

Were you running 2001 with AA and AF on ?????

 

[JZilla]

Originally posted by: Jeff7181

Originally posted by: JZilla
I think Schneider is european.

And I also got a GF4 MX420 and I have noticed no performance increase going from 30.82 to 41.09 that I am using now. I think the card is already delivering all it can.

A 1 Ghz Duron could probably push that card to it's limits.

Nah, not a Duron. I've got a XP1800 and I only get 3700 in 3dmark01, but who cares, as long as I can play SimCity. My brother (BF1942 addict) doesn't like the card, so I am trying to get him to shell out the money for a Radeon 9600.

 

[Jeff7181]

Originally posted by: JZilla

Originally posted by: Jeff7181

Originally posted by: JZilla
I think Schneider is european.

And I also got a GF4 MX420 and I have noticed no performance increase going from 30.82 to 41.09 that I am using now. I think the card is already delivering all it can.

A 1 Ghz Duron could probably push that card to it's limits.

Nah, not a Duron. I've got a XP1800 and I only get 3700 in 3dmark01, but who cares, as long as I can play SimCity. My brother (BF1942 addict) doesn't like the card, so I am trying to get him to shell out the money for a Radeon 9600.

Probably better off with a 9700 non-pro

 

[MercenaryForHire]

Originally posted by: Intelman07

They are dated at 11/5/03 so they are relatively new

Uhhh we aren't in December of 2003 yet........

Uhhh you need to learn how to read DMY military standard dates........

Downloading now, will post results later (gF4 Ti series)

- M4H

 

[nemesismk2]

Originally posted by: Jeff7181

Originally posted by: nemesismk2
Had enough time to run a few test before my lunch:-

athlon xp 1700
asrock sis746fx motherboard
xfx 128mb geforce fx5200
512mb ddr266 memory

3dmark 2001 se = 43.51=6552 44.03=6612
3dmark 03 = 43.51=1346 44.03=1375

The biggest increase with 3dmark 03 is the single texturing fillrate which increased from 465.7 MTexels/s with the 43.51 drivers to 610.4 MTexels/s with the 44.03 drivers.

The biggest loss is with 4xfsaa enabled, the 44.03 drivers result in a large performance hit compared to the 43.51 drivers. Although to be fair to Nvidia these drivers don't even list the FX5200 as being supported so I'll wait for the official drivers before doing anymore benchmarking.

Were you running 2001 with AA and AF on ?????

No the 3dmark 2001 SE scores are with AA and AF disabled, 6612 is a pretty good score for the basic FX5200 when compared to alot of the recent reviews I've seen. I saw one review with a athlon xp 2600, nforce2, ddr333 memory etc and the fx5200 only got 5090 which was less than my old gf4 mx440.

 

[Jeff7181]

That just doesn't sound right.

 

[nemesismk2]

Originally posted by: Jeff7181
That just doesn't sound right.

What doesn't sound right? The 3dmark 2001 se score of 5090 with such a powerful computer?

If that's what your talking about then I agree fully with you, if you try telling Tech Report that they willn't listen because all they say is "we only use official drivers" even if better drivers are available like the 43.51 whql drivers for example.

 

[Jeff7181]

There's gotta be some kinda compatability issue goin on with the the FX cards and 3DMark 2001 if that's all they can do. Hell... a GeForce3 Ti200 can do better than that.

 

[Regs]

The FX 5200 is like the MX version of FX's compared to its big brothers, the 5600 and the 5800 ultra. FX 5200 is pretty much a card slapped together to fit for DX9.

 

[nemesismk2]

Originally posted by: Jeff7181
There's gotta be some kinda compatability issue goin on with the the FX cards and 3DMark 2001 if that's all they can do. Hell... a GeForce3 Ti200 can do better than that.

Yeah for the most part a GF3 Ti200 is faster, the FX5200 only really beats the GF3 Ti200 with 3dmark 2003 because it supports DX9. The FX5200 was designed to replace the old gf4 mx440 and that is pretty much all it beats.

 

[OpStar]

I've seen very nice increases in games with these drivers on my FX Ultra.

Especially Splinter Cell/Unreal 2k3/BF 1942

I didn't test 3D mark as I could care less.

 

[XBoxLPU]

which FX do you have, OpStar?

 

[OpStar]

The 5800 Ultra.

 

[zsouthboy]

Any reason to move from the 30.82's on my ti4600 (besides deciding to f#ck up TV output and dvd playback, etc)?

 

[MercenaryForHire]

uNftastic!

A hefty 2000 mark boost in 3DMark2001. IQ results are damn near identical - if anything, the 44.03's have a slight edge. Performance with AA/AF on took a huge leap forward as well, and Unreal II is actually playable with the eyecandy enabled.

Verdict? Worth the upgrade, unless you're one of the poor people nVidia shafted by disabling the TV-output.

- M4H

 

[Schadenfroh]

any try these with geforce4 TI and windows xp yet? i am chicken

 

[billyjak]

I just installed them on my GF4 TI 4200 and they work great. I'm getting a little better score in 3d marks 20o1 and other benches.
I had pproblems with the 43's so I was leary to try these, but they work flawelessly.

 

[MercenaryForHire]

Originally posted by: Schadenfroh
any try these with geforce4 TI and windows xp yet? i am chicken

That's my setup. See above.

- M4H

 

[Schneider]

Well firstly i'm aussie and like i said before, the biggest performance boost is given to the FX series.
So if u have an FX gfx card, these drivers a worth a try. If ur not happy just roll back.
For me...i think i'll keep'em