*** F@H Tech Assistance Thread ***

[hertz9753]

I'm having some trouble with my 780, I started folding again a couple weeks ago and it was going fine but then a few days ago the screen wouldn't wake up. Have to reset it. Everytime I leave it for more than an hour or so the screen wont come back on and the fan is still spooled up on the 780 but the air blowing out isn't hot like when it's folding. Thought it was overclock issue, downclocked CPU/RAM, everything to stock. Same thing. Tried killing everything unimportant with task manager except the folding client, same results. Perhaps my google fu is weak, but I couldn't find this same problem anywhere. I think I might swap out memory next and see if that helps

Did you try to uninstall your Nvidia drivers? Which drivers are you using? If you are running Windows, do you have power saving turned on?

 

[hertz9753]

I'm having some trouble with my 780, I started folding again a couple weeks ago and it was going fine but then a few days ago the screen wouldn't wake up. Have to reset it. Everytime I leave it for more than an hour or so the screen wont come back on and the fan is still spooled up on the 780 but the air blowing out isn't hot like when it's folding. Thought it was overclock issue, downclocked CPU/RAM, everything to stock. Same thing. Tried killing everything unimportant with task manager except the folding client, same results. Perhaps my google fu is weak, but I couldn't find this same problem anywhere. I think I might swap out memory next and see if that helps

Are you running power saving on that rig?

 

[Irony]

I have everything power saving is turned off, sleep is turned off too. I'm on 337.8 drivers, I see theres new ones. I'll update here in a little bit

 

[mstenholm]

I have everything power saving is turned off, sleep is turned off too. I'm on 337.8 drivers, I see theres new ones. I'll update here in a little bit

In the (hidden) folder Users/you/AppData/Roaming/FAH/log you can see when your last unit stopped working plus what crashed it

 

[Irony]

Oh thanks I'll check that

Edit: I checked the logs for the last few days, they all end at like 4-5 AM with no errors or warnings, just like this:

04:43:26:WU00:FS01:0x18:*********************** Log Started 2015-01-23T04:43:26Z ***********************
04:43:26:WU00:FS01:0x18roject: 10478 (Run 0, Clone 285, Gen 8)
04:43:26:WU00:FS01:0x18:Unit: 0x0000000e538b3dba548f69ac4e8edad4
04:43:26:WU00:FS01:0x18:CPU: 0x00000000000000000000000000000000
04:43:26:WU00:FS01:0x18:Machine: 1
04:43:26:WU00:FS01:0x18igital signatures verified
04:43:26:WU00:FS01:0x18:Folding@home GPU core18
04:43:26:WU00:FS01:0x18:Version 0.0.3
04:43:26:WU00:FS01:0x18: Found a checkpoint file
04:43:46:WU00:FS01:0x18:Completed 1000000 out of 5000000 steps (20%)
04:43:46:WU00:FS01:0x18:Temperature control disabled. Requirements: single Nvidia GPU, tmax must be < 110 and twait >= 900
04:49:05:WU00:FS01:0x18:Completed 1050000 out of 5000000 steps (21%)
04:54:18:WU00:FS01:0x18:Completed 1100000 out of 5000000 steps (22%)


Lol it made smilies


Edit: I left it alone all day and it ran perfect, looks like that solved it. Thanks guys

 

[SirKeldon]

Hello team, today I was trying to set up F@H to process GPU work under MacOS X 10.8.5 with no luck. I installed the 7.4.4 client and did the common setup procedure, once identity was set up I removed CPU slot and added the GPU one, detected the GTX 770 but after many attempts no WU was assigned. Checked the log that said over and over when attempting to connect "Server work is empty" ... then the info and there I realized that CUDA wasn't available, so I deduce I was getting those errors due to this, or maybe not, my CUDA drivers are installed, activated and up to date as far as I know. I checked if it was a connection problem but with CPU slots I got work in a charm so it was out of the equation.

Checked a lil bit the web and the F@H forums and saw ppl with the same problem but not really clarifying why, just found one thread indicating to delete "GPUs.txt" to fold again, I couldn't find that file (the guy's method was for Windows) ... at this point I really don't know if GPU work is available for Mac (cause I saw other threads saying this) or is a problem of how to set up correctly the card here with some startup options for FAHClient.
If anyone can bring me some light into this i'd be very thankful!

 

[thebluebumblebee]

Hello team, today I was trying to set up F@H to process GPU work under MacOS X 10.8.5 with no luck. I installed the 7.4.4 client and did the common setup procedure, once identity was set up I removed CPU slot and added the GPU one, detected the GTX 770 but after many attempts no WU was assigned. Checked the log that said over and over when attempting to connect "Server work is empty" ... then the info and there I realized that CUDA wasn't available, so I deduce I was getting those errors due to this, or maybe not, my CUDA drivers are installed, activated and up to date as far as I know. I checked if it was a connection problem but with CPU slots I got work in a charm so it was out of the equation.

Checked a lil bit the web and the F@H forums and saw ppl with the same problem but not really clarifying why, just found one thread indicating to delete "GPUs.txt" to fold again, I couldn't find that file (the guy's method was for Windows) ... at this point I really don't know if GPU work is available for Mac (cause I saw other threads saying this) or is a problem of how to set up correctly the card here with some startup options for FAHClient.
If anyone can bring me some light into this i'd be very thankful!

I don't know anything about Mac, nor do I know of anyone on the team who does. If it was Windows, I'd tell you to reboot.

 

[revin]

Ok, so weve touched a little back when I started with the 5870, where it would stall, but now with the 7950 it's stalling in the 99.xx% atleast twice the last 3 days.
It's still giving me x17's, but I notice that when it is stalled the CPU is @ 4.8 60=/-c and only 25-35%.
Normal 1 core is 9-12% 4.6-4.8 9-12% low 40's C.
So I'm stumped as to why it's failing in the 99% range and then the CPU is jumping so high, but GPU is flatline.

I dont have a F@H log in AppData or anywhere for that matter, just wht show's in the control box.
All stock settings, but only GPU slot used, so when I quit restart it's aroung 45%, but shees with 15 min saves thats MANY hours of lost time or something else.
Thing is the log don't show an issue that far back ??????????????????

 

[mstenholm]

Ok, so weve touched a little back when I started with the 5870, where it would stall, but now with the 7950 it's stalling in the 99.xx% atleast twice the last 3 days.
It's still giving me x17's, but I notice that when it is stalled the CPU is @ 4.8 60=/-c and only 25-35%.
Normal 1 core is 9-12% 4.6-4.8 9-12% low 40's C.
So I'm stumped as to why it's failing in the 99% range and then the CPU is jumping so high, but GPU is flatline.

I dont have a F@H log in AppData or anywhere for that matter, just wht show's in the control box.
All stock settings, but only GPU slot used, so when I quit restart it's aroung 45%, but shees with 15 min saves thats MANY hours of lost time or something else.
Thing is the log don't show an issue that far back ??????????????????

You should have the log in Appdat/roamning/log if you did a standard install (win7). If the status hang at 99.9% and your log same place shows 45% the WU failed. Often pausing and re-staring will pick up from were it went wrong but if the same WU keeps failing on you then go into config/slots and delete the GPU and save and then add it again. That will dump the WU. The reason for the 99.9 could either be a bad WU or and that is more likely an too high OC.

Folding on x17s and x18s take a core from you CPU.

 

[revin]

I've looked everywhere for any F@H log, nothing found in any dir. It was a fresh install so not sure what happend.
The 45% is after I exit the F@H and restarted is what it had started at, so I believe i've been lucky and have kept the same WU at least on this last run.
Curious as why the CPU OC would be involved as it's always been the same, and I'm not using the CPU to fold just GPU Slot, it's just that somehow the F@H client is hungup ang using 30+% CPU.

All is normal after restarting the F@H client, at least for a few days lol.............. Just wasn't sure if anyone else was having issues with x17's again

 

[mx500torid]

I have the same problem with some of my AMD cards. I just restart the comp and it goes to the checkpoint and folds away. The 290x dont seem to do it but everyone of the rest did. 280x, 7950, 7850, 7770. I tried everything I could think of but nothing helped so i just live with it. I just keep afterburner up and if i see the gpu below 40c I know its not working.

 

[mstenholm]

I've looked everywhere for any F@H log, nothing found in any dir. It was a fresh install so not sure what happend.
The 45% is after I exit the F@H and restarted is what it had started at, so I believe i've been lucky and have kept the same WU at least on this last run.
Curious as why the CPU OC would be involved as it's always been the same, and I'm not using the CPU to fold just GPU Slot, it's just that somehow the F@H client is hungup ang using 30+% CPU.

All is normal after restarting the F@H client, at least for a few days lol.............. Just wasn't sure if anyone else was having issues with x17's again

I meant the GPU OC. The Appdata is a hidden folder. Sorry should have mentioned that.

 

[revin]

I'm stumped !?? No F@H log on my drive, hidden or not
What is the name of it. ie.......stanedu.log, fold.log ect..
Anyway so far so good got 62 WU's added WOOT WOOT
As long as I can keep getting any WU's finished to Dean

Thank's team,

EDIT : Any idea why my viewer dont show ? Would AB on screen affect the new core ?
ALSO WHY is my PPD down so low now, even finishing WU's early aprox 25k on the 7950 ??

 

[mstenholm]

I'm stumped !?? No F@H log on my drive, hidden or not
What is the name of it. ie.......stanedu.log, fold.log ect..
Anyway so far so good got 62 WU's added WOOT WOOT
As long as I can keep getting any WU's finished to Dean

Thank's team,

EDIT : Any idea why my viewer dont show ? Would AB on screen affect the new core ?
ALSO WHY is my PPD down so low now, even finishing WU's early aprox 25k on the 7950 ??

C:\Users\yourusername\AppData\Roaming\FAHClient log.txt

 

[thebluebumblebee]

I'm sorry, I've been watching this for several days and it only now occurred to me that HFM.net will show you your log.

 

[revin]

Finally found it ....all in ProgramData folder, below both Program Files , X86, strange but kool.....

I still never figured out how to change the HFM viewer from overclockers.net to my setup.............

EDIT Looking thru the HFM log, it has a ton of entrys not found, like those that should be in AppDat Roaming, and config's ect...
Next finished WU I'd better try to do another F@H install and see if it gets corrected.

 

[revin]

Ok seem's that 7.4.4 default Dir for Data is ProgramData, so doing a new install to differant folder.

 

[revin]

Got a fresh coustom install with choosing proper AppData folder.
Freaking amaizing thing b4 aprox 25-30k

This will get a Big boost to Dean !!

 

[Arjai]

I am having a strange issue, to me.

My 7850 works great, then stops. It finishes the job, sends it then, does it's "cleaning up" then, nothing.

I have to reboot to have it fold again by sending a new WU. I am not around, or sleeping, when this happens and then I notice it not folding, hours and hours later. Today I looked at it and it had stopped 22 hours ago!!

When I first set this up to fold, it was fine but, in the last week, or two, it has not downloaded new work, until I reboot!

Anyone have a fix for this FAH newbie? Google says...

Nada, same with Bing.

I did however learn there is a client type - advanced!!

Probably won't help me! But it looks cool!

 

[thebluebumblebee]

I am having a strange issue, to me.

My 7850 works great, then stops. It finishes the job, sends it then, does it's "cleaning up" then, nothing.

I have to reboot to have it fold again by sending a new WU. I am not around, or sleeping, when this happens and then I notice it not folding, hours and hours later. Today I looked at it and it had stopped 22 hours ago!!

When I first set this up to fold, it was fine but, in the last week, or two, it has not downloaded new work, until I reboot!

Anyone have a fix for this FAH newbie? Google says...

Nada, same with Bing.

I did however learn there is a client type - advanced!!

Probably won't help me! But it looks cool!

We need to see your log.

 

[Arjai]

I don't have a appdata/roaming/fah.log, thingy.
The current log from the Advanced View, was reset when I rebooted, to get more work.

Basically, what the Adv viewer log said...Cleaning up.

Then sat idle for 22 hours, this last time. I know my little HotSpot is iffy, sometimes. Will the Client stop trying to get work, if it can't connect, immediately?



OK, separate question. I opened a new slot for the CPU. Now my GPU is only running at 65% load.

Before, I had dropped crunching on one core, to let the Client have it.

However, even running 65% with the core running 100% the PPD indicator is jumping around between 79thou and 81thou...Perhaps 46% more?

The CPU started at 13 days, and now is at 16 hours, ETA. Is that OK, 16 hours? seems ok-ish to me. CPU temps are 95c and 83C. Kinda high but, in the safe zone, I believe.

Anybody think this is the wrong thing to do, am I pushing the e8600 too far?

 

[thebluebumblebee]

It's been a while for me, but I think you can just highlight and copy the contents of the log and then paste them here in a spoiler - to save space. I want to see what the log says from the time the client starts until after cleanup. Is there something that has changed with your system in the last week?
Do you have to reboot, or will restarting the client work?
It's our opinion that running F@H on a CPU is not worth it, especially not a dual core. You won't get 46% more. Most likely, your total PPD will fall, because your QRB for your GPU WU's will be lower. If you were to do anything, run WCG on 50% CPU's. If you're trying to run WCG, F@H CPU and F@H GPU, you'll end up getting lower PPD for both WCG and F@H. (you need to delete the CPU slot, otherwise it will grab work every time you start the client)

 

[mstenholm]

I am having a strange issue, to me.

My 7850 works great, then stops. It finishes the job, sends it then, does it's "cleaning up" then, nothing.

I have to reboot to have it fold again by sending a new WU. I am not around, or sleeping, when this happens and then I notice it not folding, hours and hours later. Today I looked at it and it had stopped 22 hours ago!!

When I first set this up to fold, it was fine but, in the last week, or two, it has not downloaded new work, until I reboot!

Anyone have a fix for this FAH newbie? Google says...

Nada, same with Bing.

I did however learn there is a client type - advanced!!

Probably won't help me! But it looks cool!

Same thing happend to me twice the past few days...nothing in the the log indicating a fault. My other GPU soldiered on both times.

 

[Arjai]

It's been a while for me, but I think you can just highlight and copy the contents of the log and then paste them here in a spoiler - to save space. I want to see what the log says from the time the client starts until after cleanup. Is there something that has changed with your system in the last week?
Do you have to reboot, or will restarting the client work?
It's our opinion that running F@H on a CPU is not worth it, especially not a dual core. You won't get 46% more. Most likely, your total PPD will fall, because your QRB for your GPU WU's will be lower. If you were to do anything, run WCG on 50% CPU's. If you're trying to run WCG, F@H CPU and F@H GPU, you'll end up getting lower PPD for both WCG and F@H. (you need to delete the CPU slot, otherwise it will grab work every time you start the client)

Thanks, PPD estimator kept falling after I made that post. Dumped the CPU slot and temps are falling and GPU is back to 95%!

If I am around, or awake, in ...7 hours 11 minutes, I will post my log. Otherwise it might be morning before I get around to it.

 

[NastyHabits]

@Arjai - Sometimes the newer WU's cause things to go haywire.
It's been awhile since I folded with my 7850, but here's what I did when things got flaky.

1. If you are overclocked, back it off to stock settings.
2. If you have recently upgraded the drivers, roll them back to the previous version.
3. Your temps are falling so that should be OK. Otherwise raise the fan speed.

Best advice came from BlueBumblebee, we need to see the logs.
Good luck. Hope it already is working by tomorrow.