HELP needed please, re: 9600 Pro

[Mr B]

Hi guys,

I'm currently working with a Sapphire Atlantis, 9600 Pro, 8x, 128MB... It's the one with the goofy heatpipe assembly tacked on it..

I'm running benchmarks, using the NF7-S system in the "sig" below (and will run them again with the IC7 rig afterwards).

I have been thru every BIOS setting I can find, or think might help (settings, 8x AGP enabled, Fast Writes enabled, 256MB aperature (where I've got a gig of DDR in the system, I've set it there...))

I installed the "Redline 1.8" program that came with the card, and Powerstrip, and enabled everything I thought might help this card out... I'm running the ATI 3.10 drivers, fresh from ATI's website. the HDD was formatted fresh for the card install, so there's no nVidia driver bits lingering about. WinXP SP1, BTW... DX9.0b.

Anyways...my problem. In the course of testing, one bench I'm running (out of several) is 3D Mark 2001SE. I wanted to run everything in default settings, as I'll be comparing this card in a future article (after the product review article first) against a FX5700 Ultra, and my trusty old 4x, DX8, Ti4200.

the thing is this, and I cannot figure this one out.... My Ti4200 gets higher scores in 2001SE than the 9600 Pro is. I'm only getting scores around 11,500 Marks, and I recall getting higher with my Ti4200 (not even overclocked..at default settings also).

What the hell am I doing wrong?!? I'd think the 9600 Pro would blow the Ti4200 away, but it's not. It does in Aquamark 3, and shadermark 2.0, but 3D Mark 2001SE, the Radeon falls on it's face....

Very perplexed.... Any and all help/suggestions would be appreciated. I know this card should wallop my Ti4200, so I've got to be missing something somewhere...

TIA!

Brian

 

[Matthias99]

NVIDIA cards generally score quite a bit higher on 3DMark01, due to heavy (although mathematically correct) optimization in their drivers. I wouldn't lose too much sleep over it (although it may be a bit late for that ).

The 9600Pro is probably about the same speed as a 4200 in straight DX6/7/8 applications with AA and AF disabled (it's certainly slower than a 4600). Try running 3DMark01 with 4XAA/8XAF and see if there's more of a difference; the 9600Pro should blow it away set like that.

 

[Mr B]

NVIDIA cards generally score quite a bit higher on 3DMark01, due to heavy (although mathematically correct) optimization in their drivers. I wouldn't lose too much sleep over it (although it may be a bit late for that ).

The 9600Pro is probably about the same speed as a 4200 in straight DX6/7/8 applications with AA and AF disabled (it's certainly slower than a 4600). Try running 3DMark01 with 4XAA/8XAF and see if there's more of a difference; the 9600Pro should blow it away set like that.

Thanks for the tips. I'll try that out, and see what happens.

How's the weather all the way down there in N. Attleboro?? Another MA native on the boards here.

Hope your Christmas was good, and Santa brought ya lots of neat stuff.

Thanks again!

Brian