Video Card Thread (Mostly Gecube x1950xt)

Yeah sure, how do I do it?

 

Sam says it's an option on the 3dmark6 - not the free version but the $20 version (pro.) I'm not in the same building with my gaming computer or I would tell you exactly.

I updated the p35 chart a little bit more with ray's memory. If you're getting 1000mhz, you must be running DDR2, is that right?

 

Yeah I have DDR2, I can run it at 1075 Mhz too, I just don't because I would have to turn down my CPU.

I'll go download the pro version and let you know. Rich, btw I'm at 3.52 Ghz now not 3.44.

 

Hey guys, just ran the fill rate test, I got 4463 and 10770. Not sure what relevance clock speed has but it was at 675/702.

 

I ran mine at the standard clock speeds, I they're 648/702...

Anyways,I ran 3dmark06 last night, and with my max overclock at 708/855 I got around 4800, which is fairly good for a simple overclock. But, then 3dmark06 randomly quite after, and I didnt have time to post a pic so I will do that all now.

 

Here is mine:

 

Damn, why are is the pic so small?

MacerM, how did you get you gpu so high? Did you give it some more juice? Do you get artifacts at that speed?

 

Wow, Ray, you really boosted your 3dmark6 score - I updated the page35 chart!

Your picture is too small because it's only 480 pixels wide - if you have a graphics program, the best width to use in this forum is 760. Any wider than that, the horizontal scroll bar appears, which is a pain. Why don't you expand it and post your picture again at 760 pixels wide.

Mike, thanks for running the fill. That's 10,700, same as Sam. The gecube rocks. Hey, I have a question for all of you guys on these fill numbers. We're all running the fill at the standard 3dmark6 resolution of 1280 x 1024, isn't that right? I am asking because I found out that the fill result is completely dependent on resolution. I found out that my x850xtpe isn't as strong as I thought. It's only bringing in about half the fill numbers that you guys are getting with the gecube, 5600, at the stock resolution. But if I drop the resolution to 1024 x 768, then I get the 8600 I was expecting. So let's make sure we are not changing the default resolution. Is that correct Sam, and Ray?

Abuzar, I posted your memory type to the chart. Hey, Sam and Mike, what kind of memory do you guys have - is Abuzar the only one with DDR2 memory?

Abuzar, yeah, let's see what you get on the fill rate test. Make sure not to change the default 1280 x 1024 resolution. Once you upgrade to the pro version of 3dmark6, you can choose your test. Uncheck the other tests, just check multi-texture fill. It only takes 10 seconds, then hit Details.

Hey, by the way, I don't know if anybody else has ever had this problem: In playing Day of Defeat today, I hooked up my brother's Apple cinema 1600 x 1024 monitor. I've used his monitor before, but never for DOD. Anyway, I couldn't get the game bright enough. The catalyst gamma or brightness settings wouldn't transfer over into the game. The brightness control for the game was maxed, and it was still too dark. I was getting killed - I couldn't see inside the buildings!. Finally I did a google, and downloaded a little 9k exe called gammasutra.exe. Then I set up some hotkeys that work inside the game. It works perfectly. -rich

 

I ran the test at 1280x1024, default settings, as I stated (I already knew the result depended on the resolution. I think thats because the score is calculated based on frame rate.)

 

I run all 3dmark06 tests at 1280x1024, or whatever the default is...

 

Each time I try to resize the photo in photoshop, it gets super blurry, which I knew would happen. I just dont understand why when I pressed "print screen" and pasted it into paint, that it pasted it at such a low resolution... My screen is running at 1280x960 @ 75 herz I believe...

 

This is what I got:
6459.542 MTexels/s

 

Ray why don't you try a couple of different things to try to figure out what is going on with your screen shots? It seems to me you should be getting images 1280 x 960 in size.

I am running my non-gaming machine at a desktop resolution of 1024 x 768. When I hit print screen it puts the screen into the windows clipboard. I paste it into paint, and also into photoshop, and save it to the desktop as a jpg. I also go right to clipboard viewer, a windows utility (w98 - I presume also xp - I run dual boot but I'm in w98 right now) and save as a jpg. All of these jpgs are 1024 x 768.

For this forum, scale your 1280 to 760 (of course keep the same aspect ratio so the height scales down also.) Does anybody have any ideas why Ray's print screen function is producing images 480 pixels wide? Wait a second Ray, don't you remember that one time when you posted that giant screen shot! How did you get that one so big?

Abuzar, you get about 6500 fill. That is about what I thought (I was thinking 6900 from charts I had seen - that's more or less the same thing.) So you're basically getting 2/3 the fill as the gecube or as Sam's card, for $85, versus the $670 that Sam spent 1 1/2 years ago, and the $500 that I spent in 2005 for my 5600 fill. Your 3dmark6 score is great - Mike would probably say you're a little overbalanced with stronger cpu than graphics card.

I guess if we're willing to put up with a little bit of pain, (maybe a lot of pain) the Gecube is not a bad deal right now for $230.

 

Yeah my CPU probably is a bit too powerful. Anyway I bet you I could beat your frame rates in games even when you have that nicer card, if you guys are using a P4.

I might get an 8800 GTX later on...

 

I bet you're right! Let's not wait for me though!

Allright, a new challenge! P4 and faster graphics card versus core 2 duo and slower card!

Ray, we need you!

Hey, I bet Ray is up for that. You can AIM him at his user name. He and Travis got on the same BF2 server a while back and got the same framerates testing omega drivers versus stock catalyst. Since they have the same cpu and same graphics card and almost same 3dmark6 (until Ray just boosted his numbers with some major oc'ing - he was about 4426 before I think) it was a perfect test of the two drivers. They came out about the same, which is what I expected, but kind of a puzzle since for Ray the omega drivers doubled his fps. (There must have been some problem with his catalyst installation.)

Hey Ray, how about meeting up with Abuzar on BF2, and let's see how your faster gecube (10000 fill versus his 6500 fill) compares with his much faster cpu (core 2 duo - faster than Sam's) versus your P4. That will be a great test to see, at least for BF2, how important is the graphics card versus the cpu! If you meet up you can make sure you're both running the same resolution and same settings - don't kill each other before you get the fps numbers

Abuzar, do you have BF2? If not, you can get the free demo. Maybe Ray (who has the whole game) will be willing to meet up with you on a demo server.

THE CHALLENGE IS ON: THE NEW MATCHUP!

 

Sure I'll go for it. How about we post what we get on the HDR and SM2 (or SM3) benches from 3dmark06... Those are where the graphics cards are tested.

 

SM 2.0 Score 1999 Marks
SM 3.0 Score 2180 Marks

Yeah sure I'm up for it. I don't have the game though so I will have to use the demo.

 

Oh, not sure if the demo is multiplayer though...

SM2 Score: 2017
HDR/SM3 Score: 2346

Btw guys, I finaly got the print screen thing to work, so I will post a screen shot of 3dmark06 score up here in the next 10-15 minutes.

EDIT: Just remembered I had to do something for school, so I will have to postpone the 3dmark thing till tomarrow.

 

Haha I have three exams tomorrow so I know what that's like. In fact I should go back to studying...

 

Rich, don't forget that despite more RAM and the PCI express architecture, I have an older chip design in my card, so a lower fill rate is probably to be expected of me and abuzar's X1900GT. There's got to be something in the number 5 of the X1950.
I have elected to run my card in overclocked mode for a few games now, rather than just for an experiment. Now I've finally got top-quality case fans, I can turn them up beyond 450rpm (read around 900) and my PC is still almost the quietest air cooled PC at a LAN party of 45 or so people. The only reason it's noisy is my WD Raptor. The reason why I'll be running my GPU at its overclocked speed is because I found out I can increase my resolution from 1680x1050 to 1920x1200 in the UT3 demo and get the same frame rate as before. That's quite a big jump, and will come in handy for the big games coming out later on.

With your little 3dmark race, the underdog should really be the Core 2 duo system, since 3dmark cares far more about graphics than your CPU. Ray's score is good evidence of that.

I'm happy at the moment. Despite having a bit of a USB hiccup shortly prior to the event, my PC ran beautifully through the tournaments at FragSoc, and I even came close to winning one of them (UT2004 Instagib). I may have had to be awake for 39 hours with only two hours sleep in the middle, but it was still probably one of the top 10 weekends I've had. I've been to 9 fragsoc events prior, but that's got to be one of the best.
One of the interesting things I noted is that about 5 people now run the Thermalright HR-03 & Ultra 120 combo that I have. The upside of this is that most of the PCs at the event make hardly any noise! Evidently they're really catching on.