Just received my X800 XT from Zipzoomfly!!!!

[GeneralGrievous]

I'm gonna install it now and do some benchmarking before I overclock. If there are any questions I can answer please post them here.

 

[Pete]

OMG PICS!!11oneone

I mean, benchmarks!

I'd also like to hear your opinion of its texture filtering (compared to your previous card). Do you notice mipmap bands or excessive shimmer?

 

[Corporate Thug]

does it look just like the reference model???

 

[tbaron55]

Is this the Abit one that was listed for about two hours? I ordered one of those about 2 mins after the first post about it and I received email stating it wouldn't arrive until June 28...

You must have ordered the split second it came online???

Thanks,

Tim B.

 

[GeneralGrievous]

Ya, I did get it literally the second it came out. And shady, it looks like the reference model.

Just did some quick Farcry benchmarking near the indoor part of the first level.

At 1600x1200 2xAA 16x performance AF:
Minimum FPS: 51
Average FPS: 64

Upping to 4xAA
Minimum FPS: 42
Average FPS: 55

I wasn't comfortable with this so I set it back to 2xAA. For everything I forced it in the drivers (app preference unchecked). If there is a better way of doing this tell me.

I popped my 9800 Pro (440/360) back in and started running it at the 2xAA settings.
Minimum FPS: 23
Average FPS: 30

Here are some pics from the first checkpoint

This is the 9800 Pro running Farcry at the above settings.

This is the X800XT.

I probably should have taken the screens at the checkpoint but I didn't like the shots so I moved up and tried to duplicate it a second time.

Looking at them the texture at the side of the hut looks slightly better on the 9800. Decide for yourself.

 

[Avalon]

Awesome. Thanks for deciding to share with us. I await the beauty that will be your results.

 

[imported_x1]

hmm... looks the same to me. These IQ samples reviewers post always boggle my mind, how can they tell the difference , maybe I'm just blind. Anyway, for that speed increase I'd definately accept a 5-10% drop in IQ.

 

[Avalon]

Wouldn't it be better to just set your x800 panel to application preference and assign the AA and AF values in FarCry itself?

 

[edmundoab]

very fast on the benchmark. Care to share the score obtained from 3dmark 2003?

 

[GeneralGrievous]

I tried that but for some reason Farcry wouldn't run over Medium texture filtering.

UT2k4 benchmark: Deathmatch, Ice Tomb, 15 bots. 1600x1200 4x/16x
Minimum FPS: 58
Average FPS: 76

6xAA really crippled the card into the mid 40s.

This game really isn't the best for the card, as I could run 1600x1200 2x/16x with a 9800 Pro and obtain similar minimum results.

 

[Pete]

GG, do you notice a difference b/w 4xAF set in-game and 16xAF set via CP? Maybe you could try those registry settings ComputerBase found to see what disabling trylinear does for quality and framerate.

 

[GeneralGrievous]

This is 4x enabled in game.

This is 16x enabled in CP.

Note the bottom of the tree. Perhaps its because I am closer on the 2nd shot, but it does look blurred out a bit on the first. There are other differences as well.

ATI has an excellent AF algorithm anyway, the performance difference even on older cards was very small.

 

[Corporate Thug]

very promising benchies!!!

 

[nRollo]

Originally posted by: Pete
GG, do you notice a difference b/w 4xAF set in-game and 16xAF set via CP? Maybe you could try those registry settings ComputerBase found to see what disabling trylinear does for quality and framerate.

If I had this card, I would disable the brilinear and play it with the best IQ possible. The framerates are still going to be great.

Brilinear make sense for the 9600XT, it makes no sense for the X800.

 

[GeneralGrievous]

I don't have 3dmark03 but I'll download it tonight.

But here's an idea of what the card gets at stock settings.
Here's stock cooling overclocked settings.

 

[Gagabiji]

Originally posted by: x1
hmm... looks the same to me. These IQ samples reviewers post always boggle my mind, how can they tell the difference , maybe I'm just blind. Anyway, for that speed increase I'd definately accept a 5-10% drop in IQ.

Same here. Unless it's apples to oranges I can't tell the difference.

 

[nRollo]

Originally posted by: GeneralGrievous
This is 4x enabled in game.

This is 16x enabled in CP.

Note the bottom of the tree. Perhaps its because I am closer on the 2nd shot, but it does look blurred out a bit on the first. There are other differences as well.

ATI has an excellent AF algorithm anyway, the performance difference even on older cards was very small.

Kudos for being honest about the IQ differences General Grievous. You could have just said "There's none I can see" and we'd have to take you at your word or buy X800s.

 

[GeneralGrievous]

Kudos for being honest about the IQ differences General Grievous. You could have just said "There's none I can see" and we'd have to take you at your word or buy X800s.

Those shots are simply comparing 4x AF to 16x AF Rollo. Of course there will be an IQ difference.

If you want to see 9800 vs X800 shots check my 2nd post in this thread. If you can find a significant difference due to ATI's new cheat please show it to me.

On another note, to be honest I am a bit disappointed in UT2K4 performance. Enabling 4XAA was nice but I did expect 6XAA to be playable even if there is no real visible difference. I did have a significantly higher average FPS on the X800 but the minimum was the same. I think this is CPU limitation here, with an A64 I would probably have higher min FPS on the X800.

 

[SilentRunning]

Originally posted by: SirDude

Originally posted by: x1
hmm... looks the same to me. These IQ samples reviewers post always boggle my mind, how can they tell the difference , maybe I'm just blind. Anyway, for that speed increase I'd definately accept a 5-10% drop in IQ.

Same here. Unless it's apples to oranges I can't tell the difference.

Maybe its just me, but the jpeg compression artifacts kind of make it hard to really compare the images.

 

[Avalon]

I can't tell the difference between the two trees. I guess I just don't notice the effects of AF. Good thing, because then I don't have to enable it and take the performance hit

 

[Compddd]

Since my 9700 Pro days and onward I've always ran all my games 4xAA and 16xAF and always got good performance? Thats why I'm looking forward to the X800XT PE so much

 

[MrPabulum]

Those numbers for Farcry look great. I read in this month's MaximumPC that enabling the AA within Farcry produced better results, since Crytek implemented AA more efficiently. Odd that medium multisampling wouldn't work for you. . Maybe an upgrade to the Cat 4.6s would help?

 

[Insomniak]

What did you upgrade from? If it was newer than a GeForce 4, you wasted money :-\

 

[GeneralGrievous]

What did you upgrade from? If it was newer than a GeForce 4, you wasted money :-

I upgraded from a 9800 Pro. The $250 was well worth it over the next 2 years. The geforce 4 is a piece of garbage.

Maybe an upgrade to the Cat 4.6s would help?

I'm actually already using the cat 4.6s. They are good drivers, my brother got a 350 pt increase in 3dmark03 on his.

Also 1600x1200 at 4x/16x is quite a bit to ask from a card. It's certainly playable. That's a good point though, I will try setting CP at app preference and enabling AA in game.

All things considered though, I am getting significantly better results than anand did in his tests.

Since my 9700 Pro days and onward I've always ran all my games 4xAA and 16xAF and always got good performance?

What games are you playing? Counterstrike? And what resolution?

 

[Corporate Thug]

Originally posted by: Insomniak
What did you upgrade from? If it was newer than a GeForce 4, you wasted money :-\

why do you say that? i couldnt run farcry worth a damn on a 4600 and i barely run it with a 5900 @ 5950 @ 1280*768 no AA/AF