EUE Exceded limit

[PorPorMe]

FahSpy is telling me that I have exceeded the EUE limit and the GPU client is pausing for 24 hours. What does this mean? and why?

[ThunderRd]

EUE means that the WU has ended early; the computations haven't finished and the WU is now trashed. This could be heat, an unstable client setup, or some other stability problem. If you get enough EUEs in sucession, the client will stop processing for 24 hours. What this accomplishes is beyond me, but I would guess that Stanford doesn't want the same faulty client erroring out ad infinitum, so it gives the user a chance to correct what's wrong.

You will have to check out your entire setup and take a look at the temps carefully. Only you can determine what is causing the WUs to fail, but from what I can tell from the experience of other users it's usually a heat problem.

[PorPorMe]

Thanks, I think it must be the IDE drive I've been messing with. I ran Everests stress test and it shut down because of HD temps even though I have a fan blowing right between them. It surprised me as all my temps are looking better than ever. I'll pull that drive and run the test agian. will let you know.

[PorPorMe]

Originally Posted by ThunderRd

EUE means that the WU has ended early; the computations haven't finished and the WU is now trashed. This could be heat, an unstable client setup, or some other stability problem. If you get enough EUEs in sucession, the client will stop processing for 24 hours. What this accomplishes is beyond me, but I would guess that Stanford doesn't want the same faulty client erroring out ad infinitum, so it gives the user a chance to correct what's wrong.

You will have to check out your entire setup and take a look at the temps carefully. Only you can determine what is causing the WUs to fail, but from what I can tell from the experience of other users it's usually a heat problem.

It appears heat was the issue but it baffles me how or why.
I've been trying to resurrect an IDE drive so It's been connected for over a couple of weeks w/o any problems. I ran Everest's stress test and got a test shut down because of HD over heating. I never gave it a second thought about the GPU client. when you mentioned heat
I got around to disconnecting it and then ran OCCT. To much was happening around here so the only screenshot of that time I was able to get is The OCCT. The others I've take sinse as this is the way I'm running.
If you have a suggestion-feel free. but she feels smooth and strong.

[PorPorMe]

Something else is causing this. I just got the same notice in FahSpy and then looked at the log. I hope it makes sense to you because my temps are great and she is stable. GPU has stopped.

EDIT:
I'm going back to stock for now.

[ThunderRd]

Is the card overclocked? Post the output of GPU-z if you have it.

[PorPorMe]

It was OC'd but all i have is this.

[ThunderRd]

There seems to be something wrong with Precision. You shader clock is set to 2000 but is only working at 602. Do you know why? Was the client running when that screenshot was taken, because 41 C seems almost impossible at load.

I'd set all the card clocks back to stock, and see if the EUEs go away. If they do, increase the shaders slowly until the WUs fail again. The other clock settings have very little effect on folding, and will increase the heat output. The shaders are the only one worth playing with.

[PorPorMe]

The client was not running at the time, just OC. I have sinse gone stock. Video was sluggish- removed Precision and changed drivers to 185.15.video still sluggish.
Nvidia control is all st to default. I moved the pageig file to another partition and enlarged it.
Could this have anything to do with memory situation in Hardware discussion?
I'm going to run memtest next and then another full virus, maleware-you name it scan-should keep my Pet busy for the day!

[PorPorMe]

It appears that i have fixed the graphics problem and a little time will tell with the GPU client. Memtest went great. noob gave me some clues to look at as far as memory is concerned. Loading Xp back on the laptop so I can read both the manual and my bios at the same time. I want to get this right before doing any more. I think I have the page file's set better but want to ck that to. I'm dotting my i's and checking any thing on this one.

[ThunderRd]

You changed so many things at one time that it will be impossible to tell which one was teh culprit, so I am now lost. The only thing you can do now is cross your fingers along with your t's.

Did the client run properly after you changed to stock clocks and did nothing else? This is what I needed to know.

Good luck anyhow.

[PorPorMe]

I can't tell you that. I just went through the logs and got to the last shut down but as you said I've made a number of changes and I can't pinpoint each one.
I believe it was Precision that started this-even going back to stock, I just used it for the fan, I still had problems.I just can't say for sure about the GPU client because of the time lag, but sinse I've deleted it the client has been running ok. Best I can give you.

[PorPorMe]

The GPU client just quit! After uploading the last EUE it says I've exceded my EUE limit and it's shutting down for 24 hours.
What's up with that?

[Daniel ~]

For 24 hours? Weird!

[ThunderRd]

It's normal, actually it is a feature. Stanford doesn't want the servers pelted with errored-out partial-credit WUs, so it limits the number of consecutive EUEs. The 24-hour bit is because many donors aren't at their machines all the time. If a setup goes bad for some reason, and it's on a Friday evening for example, and the donor won't have physical access until Monday morning, a ton of bad data will be uploaded until the donor straightens out the situation. It's not good for anyone, and so the client shuts down and tries again in 24 hours.

The results from EUEs aren't worth a whole lot to the science; the data are most often questionable or bad anyhow. They give you partial credit more out of courtesy than anything else, to compensate you for your cycles.

Your box has a problem if you are getting so many EUEs. They aren't normal.

What you should be looking at is why you are having so many bad results. Are you having a heat problem? Is the card overclocked?

[PorPorMe]

The simple answer is all of the above-thanks, I'll make some changes.

[gaulty]

hi all...i have 2 identical machines which i use as part of my folding farm, identical in everyway except the os, the xp machine has never stopped once for eue limit, the vista machine stops every other day, both show same core temperatures...maybe windows 7 can't come quick enough...bye bye vista

[ThunderRd]

I'm still using XP for GPU folding until I'm ready to make the jump to Win7. Unless the Linux GPU client comes first...then it's bye bye Microsoft....