• Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!

Strange artifacting... X850 - HELP

Overclockers is supported by our readers. When you click a link to make a purchase, we may earn a commission. Learn More.

RobxMcCarthy

Member
Joined
Jul 18, 2004
Ok, there has to be SOMETHING I'm missing, this is driving me nuts..

I'm overclocking my X850, all the timings are set properly. The TRD was even set high (6). And I can't get it artifact free when "scanning" above 550 ram. Even with the 12 pipeline bios. (it infact can not run the 16 pipeline bios as the ram freaks out, even when underclocked I have to bring it to 480 in order to get it stable).

But here's the clincher!

I can run 670 ram while running the spinning cube with NO artifacts, none.. I can actually go up to 680 before I get the normal ram model spikes. When I switch to "scan for artifacts" the model appears to be "spattered" with yellow paint, repeatedly.. That's The ONLY artifact i ever get, same as the core artifact...in fact. Which is weird; usually ram and core artifacts are different (in my experience.)

So can ANYONE help me out, I CAN'T figure out this problem and it's driving me insane. The ram on the card remains relatively "warm" to the touch when running 660 on the spinning cube for 30 minutes.. Yet artifacts at 555 when "scanning" As well as in games (rarely).

Does anyone know anything that could cause this strange kind of artifacting?
 
I hate to sound like a douchebag, but did you ever consider the fact that you have your card soft-modded? Even with 12 pipelines in the softmod, it's possible the mod itself could be causing such artifacts. And please don't give me a reply until you are 100% certain it works with no artifacts whatever at stock. I'm guessing it's a) the soft-mod itself, b) the overlcocks, c) the pipelines, or d) the temps. Sorry to break it to you, but one of the four must be screwed up. The only other explanation is that, if it performed such things at stock, the core might be screwed.
 
It worked the same way BEFORE I soft modded.

Then I soft modded to 16 pipes and couldn't even run it at 500 500. So I backed it up to the original bios.

My temps are 49 core LOAD with cool to the touch memory... Not temps, not soft mod.

The overclocks aren't high at all. My absolute max ram before any artifacts get scanned is 542 (3 hour scan).

therefore it's not A B C or D... Hate to eound like a douche-bag, but I think that the chips might be screwed up or there might be a voltage dip..

Also did you read that it hits 680?
 
I would agree with the softmod/oc scapegoating but I am cosntantly hearing complaints of this exact problem in corelation to the 1.6 driver update for the card. Check your driver versions and scan forums for any reported bugs.

It sounds like your cooling is fine so I'd pass on blaming the heat.
 
Well, what do you mean by softmod scapegoating? When I first got the card (before doing any soft modding) I was only able to hit 542 ram, or thereabouts.

As far as 1.6 driver revision, not quite sure what you mean here either. Do you mean the 1.6 bios? In this case I'm not quite sure where to find a non 1.6 bios but I believe that is indeed the revision used. But as of now I'm using the stock bios with Omega drivers - 6.1 cat. I can try different drivers, however, I'm not sure if that will effect the problem


As far as the percision goes, I was indeed using high, however The artifacting was very visible; yet unlike normal ram artifacting. I'll grab a shot of it later on to show you what I mean.
 
No, it looks like random yellow dots. And when I hit about 650 or so it resembles yellow paint being thrown at a wall repeatedly. It's strange looking.
 
In game I do indeed get very moderate model spiking and texture artifacting after about 555. 620+ causes games to crash after a few minutes. But the game will inevidably crash at anything at 555 or above.
 
No ideas? I'm running out of things to do aside from volt modding. I'd like to use that as a last resort.

Are there any other timing alterations from the base set for the PRO that I could change? My TRP is already at 6.
 
Seems to me like you got a bum IC or something...thats when it only goes crazy once the card gets really stressed. I'd tighten back to 5, and voltmod it. Topping out at about 550 is rare, but it does happen. After volt mods, you should be able to get around 580 clean, I'd estimate.
 
Yeah, but what's weird is shown in the pic. The card will run the rotating cube without artifacts at 665 memory (shown). I mainly just want to be able to enable 16 pipes, which hasn't happened as of yet.
 
I've noticed a lotta wierd things with atitool also... for example, with my gto atitool crashes at about 520mhz when i do find max mem, but it's stable up until 560mhz when I do 3dmark loops...
 
after i got condensation between my gpu and pcb thats just what happend to my x800xtpe..... luckily after drying out it has come back to normal...
 
I think there might be something wrong with my board...

I ran 3dmark (no visible artifacts) at 590 memory.. However it crashed during one of the fill tests and corrupted the drivers....the F? This is really confusing me, as I've never seen a video card respond like this one. I almost want to RMA it again, as I'm worried that the card will degrade further in the future.
 
g0dM@n said:
RMA again? what happened with the first one...

The first one just stopped working one day. He had it out of the system to replace the water in the cooling loop. When t went back in, it gave the "the power connector isnt connected" error. And it certainly was connected.

Then his water system poored water all over his 9800. So its "lucky" that the card had a problem, because it was saved from a definite un RMA-able failure due to lots and lots of water :- ]
 
Back