• We’re currently investigating an issue related to the forum theme and styling that is impacting page layout and visual formatting. The problem has been identified, and we are actively working on a resolution. There is no impact to user data or functionality, this is strictly a front-end display issue. We’ll post an update once the fix has been deployed. Thanks for your patience while we get this sorted.

________________________overclocking x800xl________________________

mdchesne

Banned
i overclock my x800xl using ati tray tools.

my question is this:

do i really trust the program to find the best core and mem speed using their user-friendly test?
and how long do i test for artifacts before i know it's stable?
(i ran it for 30 seconds before running into an artifact, long enough?)
 
This is a new tool (never heard of it before).

ATItool has been around for quite some time. Many have experience with it and can help you if you have problems/questions.
 
ok, so now i can only achieve a 420 core and 508 mem speed before it picks up artifacts within 10 seconds after fiddling around with speeds. but i run the test artifact right after i boot using my 436/520 i have now and i get nothing for 20-30 seconds. then it picks some up, goes into a bunch of 1-second "found artifact!" results, then runs another 20seconds before finding an artifact. Then, testing on ati tray tools, i get about 100 seconds before artifacts are discovered.


how long is long enough for a good stable card? 30 sec? 5 min?
 
Originally posted by: mdchesne


how long is long enough for a good stable card? 30 sec? 5 min?

Never. Artifacts shouldn't be there at all. Lower your clock speeds until no artifacts appear.

I don't know how that program works, I usually just overclock, play a game for a while. If artifacts appear, I lower it. I repeat the process until no artifacts exist and gameplay is fine. I consider that a stable GPU overclock.
 
Originally posted by: mdchesne
ok, so now i can only achieve a 420 core and 508 mem speed before it picks up artifacts within 10 seconds after fiddling around with speeds. but i run the test artifact right after i boot using my 436/520 i have now and i get nothing for 20-30 seconds. then it picks some up, goes into a bunch of 1-second "found artifact!" results, then runs another 20seconds before finding an artifact. how long is long enough for a good stable card? 30 sec? 5 min?

That would be enough if you plan to use the computer for 5 minutes only! If you plan to play a game or watch a video for 2 hours, it had better run for that long with no artifacts. Only then, would I call it "stable".

You may need to improve the cooling of your case. That should let you overclcok more and still be stable. If you need more, you can improve the cooling of your graphics card. But, that may require a cooler upgrade, which would void the warranty.
 
ah, i see.

ok, well using atitool now, i can't get anything better than 410/1010 (410:506) without artifacts. anyone know why this could be so? powercolor's are supposed to overclock very well and it's larger than usual HSF should be keeping the temp down enough to enable further overclocking. and i dont think the lian-li pc-60 has coolign issues.

anyone have any suggestions?
 
ATItool stresses the GPU, which increases the GPU temperature like it would happen when you put stress on GPU in real life (playing games or video). While it is doing that, it searches for artifacts. If it detects artifacts, it reduces your GPU and memory frequency. It keeps doing that until it finds the highest clock rates that are stable (under load).

After that, you may still need to reduce your clock rates if you see artifacts in a game.

As I said before, you can improve the performance of your graphics card by improving cooling.
 
i've actually been manually messing with the speeds, then testing for artfacts for about 5 minutes after each 5mhz change. i guess the automatic way is easier (thought it might be flawed so i didn't trust it)
 
What I did, when I had an ATI card, was to use the ATItool automatic overclock optimizer to get me close to where the max was. Then, I would do what you describe to fine tune my overclock rates.
 
using the atitool, i got 426:504 whcih is was running stable at 70C load for half an hour. no artifacts or warnings. guess i do need to get an arctic cooling silencer to get that heat down. i'm still wondering how review sites get a 455:550 overclock without changing the stock cooling. must be running with a 60-fan case.
 
Originally posted by: Actaeon
Originally posted by: mdchesne


how long is long enough for a good stable card? 30 sec? 5 min?

Never. Artifacts shouldn't be there at all. Lower your clock speeds until no artifacts appear.

I don't know how that program works, I usually just overclock, play a game for a while. If artifacts appear, I lower it. I repeat the process until no artifacts exist and gameplay is fine. I consider that a stable GPU overclock.

Here's my theory on how it works. The program does a screenshot of the window (screenshots show artifacts), at default speeds. Then, for every sequential overclock, it compares the stock screenshot with the current screenshot. If the current screenshot shows artifacts (differs at all from stock screenshot), it alerts you.

Yes, you can absoutely trust it to find artifacts.
 
alright, i got it stable, no artifacts. now is there a way for atitools to load those settings each reboot? because everytime i reboot it goes back to stock even after saving the speeds in the atitools profile database.
 
Originally posted by: xtknight
Originally posted by: Actaeon
Originally posted by: mdchesne


how long is long enough for a good stable card? 30 sec? 5 min?

Never. Artifacts shouldn't be there at all. Lower your clock speeds until no artifacts appear.

I don't know how that program works, I usually just overclock, play a game for a while. If artifacts appear, I lower it. I repeat the process until no artifacts exist and gameplay is fine. I consider that a stable GPU overclock.

Here's my theory on how it works. The program does a screenshot of the window (screenshots show artifacts), at default speeds. Then, for every sequential overclock, it compares the stock screenshot with the current screenshot. If the current screenshot shows artifacts (differs at all from stock screenshot), it alerts you.

Yes, you can absoutely trust it to find artifacts.

I didn't say you can't trust it, I am sure you could. I just didn't know how it worked. I just explained my testing method.

Whatever floats your boat.
 
Originally posted by: mdchesne
using the atitool, i got 426:504 whcih is was running stable at 70C load for half an hour. no artifacts or warnings. guess i do need to get an arctic cooling silencer to get that heat down. i'm still wondering how review sites get a 455:550 overclock without changing the stock cooling. must be running with a 60-fan case.

How is your case temp? the lower it is natually, the lower CPU/GPU temps will be.

 
system: 35C
cpu: 33C
aux: 17C (I don;t know what this is for, but wish the other two temps were like this)

idle, of course. i'll run prime 95 for a day and list the load later
 
Back
Top