"Artifacts" from overclocking core? (9800Pro)

[Cherub]

I have an ATi-branded 9800Pro. At 450 MHz core, both 3DMark2001SE and 3DMark03 run stably, but two tests, Dragothic and Mother Nature show extreme "spakling". It starts around 425 MHz and gets gradually worse. It does not look like memory artifacts, and does not appear in any other tests. Also, in Dragothic it only appears on the dragon's back and in Mother Nature it appears mostly in the water and on some other textures. Is this normal? Is it a sign that the card is not overclocking well, or is it normal?
can this be remedied by use of a different driver?

Any help/insights appreciated.

 

[Don66]

Try lowering your O/C abit.
Sparkling isn't normal.

 

[Corporate Thug]

450 mhz is really pushing the core, esp if it is stock cooling

 

[Cherub]

Don66 -

When I asked if it was normal, I should have asked, "Is it common when overclocking." Obviously it's not normal functioning. I've just never used a Radeon model before, and I've never seen this with an nVidia card before. Generally, GF4's start skipping or pausing when the core is pushed too far. Also, it's a little bizarre that only certain parts of certain tests show the issue. It suggests that only a certain operation is having a problem.

shady06 -

I realize that 450 is pretty high, but the stability is 100% even beyond that. Also, the sparkling starts much lower, around 420-425, and gradually increases. I've never seen a blue screen or lock.

Thanks both of you for the responses.

 

[nick1985]

Originally posted by: Don66
Try lowering your O/C abit.
Sparkling isn't normal.

spakling, not sparking

**EDIT** he said "spakling" in his first post, then "sparkling" in his second.

 

[Cherub]

nick1985 -

Trust me, if I really saw spakling on my monitor, I'd get a new one. And I'd make the maintenance guy pay for it!

 

[high]

Hey I second the poster's problem. I get this sparkling on the same core with my 9800np. 100% Stable at 450+ but I get the sparkling around 420 aswell. It's due to pixel shader but I don't like it so my core is backed to 420 . Unfortunately I have not found the answer anywhere but it would be good to know.

 

[high]

Bump - anyone have sparkling problems with ATI's?

 

[Robor]

It's a result of a core overclock too high. Mine does the same thing if I go past 410. It's most noticeable in the flowing water during the Mother Nature test as well as the rear right leg of the small rhino on the Pixel Shader 2.0 test. The higher the core overclock the worse the "sparkles" get. Just back it down until you don't see any more sparkles and then maybe drop it one more notch just to be safe.

 

[Jeff7181]

You mean it doesn't just sparkle when it's new and shiny?

 

[Vonkhan]

sparkling = overheating

 

[high]

Overheating? How come you can go from 420--->450+ with no artifacting just slight sparkling that gradually gets worse. I just think it's certain operations that the core cannot handle @ that speed, personally. Sucks eitherw ay because it has soo much more headroom.

 

[Matthias99]

I don't know about other people, but I'd call random white dots all over the place "artifacting".

Obviously, *part* of the core is still functional at those speeds (mine will go all the way up to about 500 without major issues), but part of it isn't. What this does mean is that improved cooling will keep it running longer -- on overclockers.com, a number of 9800Pros with water cooling are listed as running at well over 475 core without artifacts. YMMV.

 

[Rogodin2]

I'd call it shimmering white dots, all of my oc's exibit this when pushed too hard-I use the ATI fur demo to quickly see if my core and mem are too high-been able to get my 9800np (stock cooling) to 452/411 before the "shimmering" kicks in.

It's common among oc freaks.

rogo

 

[Cherub]

Overheating? How come you can go from 420--->450+ with no artifacting just slight sparkling that gradually gets worse. I just think it's certain operations that the core cannot handle @ that speed, personally. Sucks eitherw ay because it has soo much more headroom.

Agreed.

These sparkles are not artifacts in the common sense of memory artifacts and certainly look different, but to the extent that they are graphic anomalies caused by a malfunction of some part of the video subsystem they could certainly be called "artifacts". The real difference is that memory artifacts are randomly placed and sized while the ones being discussed here are not.

The point of my question is to determine whether this is a common experience when overclocking, or an oddity of my particular card. I do have watercooling, but I am trying to determine whether I will keep this particular 9800Pro BEFORE modifying it with watercooling.

 

[high]

What kind of cooling are you using if you're able to hit 450 WITHOUT sparkling??? VERY nice oc either way.

 

[Whitedog]

I don't get why you have the need to OC a 9800Pro...?

Is it just so you'll get higher scores on 3DMark??

 

[Evdawg]

I don't get why you have the need to OC a 9800Pro...?
Is it just so you'll get higher scores on 3DMark??

Thats exACTLY why dog.

EDIT: agh screwed up the quote =(

 

[Lonyo]

Originally posted by: Whitedog
I don't get why you have the need to OC a 9800Pro...?

Is it just so you'll get higher scores on 3DMark??

Yeah, it's so stupid to get a bit of extra speed - for free. God, you're all such fools for getting extra speed for free, why bother? I mean, it's free. How stupid of you.

 

[Cherub]

Every little bit of speed counts. It may be the difference between 4x AA and 6x AA. Life altering? No. But it is free and besides overclocking is a hobby. Why does anyone modify a car so that it makes more horsepower or handles better; you don't need to go faster than the speed limit, right? Making improvements by effort and talent (or luck) is satisfying.

More importantly, overclocking a card increases its sustainability a little bit as new games push it harder than those that were out when it was top of the line.

 

[Bartokomus]

Isn't there really two kinds of o/c-ers? the kind who do it as a pure hobby, for the joy of making there machine go !vroom! and then the kind who start to o/c once there components are no longer optimal.
I fall into the second category for the most part; i o/c once i feel that my H/W is holding me back, as that usually means it's nearing end of life with me...

 

[Bartokomus]

Isn't there really two kinds of o/c-ers? the kind who do it as a pure hobby, for the joy of making there machine go !vroom! and then the kind who start to o/c once there components are no longer optimal.
I fall into the second category for the most part; i o/c once i feel that my H/W is holding me back, as that usually means it's nearing end of life with me...

 

[Cherub]

We're pretty far off topic here, but what the heck...

Consider that the 9800Pro can be had for $299 and the 9800XT goes for $499 at Best Buy. If I can overclock my 9800Pro to XT speeds, I just saved $200, in theory. Of course, this ignores the fact that the XT can also be overclocked from stock speeds and had additional features such as thermal monitoring, but it still proves the point behind the mentality of much of the overclocking that occurs.