F@H error - will this resolve itself?

[.far]

Hi,
Just started with F2H for the challenge, and have got an error. The card has been running for a few hours (and probably half a year 24/7 prior to this on Boinc projects without issue).

Is this something I have to resolve or something that it will sort out? The GPU slot does appear unresponsive even though the log appears to say something about cleaning up..

Thanks for any guidance/advice,
Far

-----------------------------------
02:52:13:WU01:FS01:0x21roject: 13201 (Run 3, Clone 5, Gen 72)
02:52:13:WU01:FS01:0x21:Unit: 0x0000003dab436c66577fedfe1de59c47
02:52:13:WU01:FS01:0x21:CPU: 0x00000000000000000000000000000000
02:52:13:WU01:FS01:0x21:Machine: 1
02:52:13:WU01:FS01:0x21:Reading tar file core.xml
02:52:13:WU01:FS01:0x21:Reading tar file integrator.xml
02:52:13:WU01:FS01:0x21:Reading tar file state.xml
02:52:14:WU01:FS01:0x21:Reading tar file system.xml
02:52:16:WU01:FS01:0x21igital signatures verified
02:52:16:WU01:FS01:0x21:Folding@home GPU Core21 Folding@home Core
02:52:16:WU01:FS01:0x21:Version 0.0.17
02:52:23:WU02:FS00:0xa7:Completed 17600 out of 80000 steps (22%)
02:53:25:WU02:FS00:0xa7:Completed 18400 out of 80000 steps (23%)
02:53:45:WU01:FS01:0x21:ERROR:exception: Error downloading array interactionCount: clEnqueueReadBuffer (-5)
02:53:45:WU01:FS01:0x21:Saving result file logfile_01.txt
02:53:45:WU01:FS01:0x21:Saving result file log.txt
02:53:45:WU01:FS01:0x21:Folding@home Core Shutdown: BAD_WORK_UNIT
02:53:48:WARNING:WU01:FS01:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
02:53:48:WU01:FS01:Sending unit results: id:01 state:SEND error:FAULTY project:13201 run:3 clone:5 gen:72 core:0x21 unit:0x0000003dab436c66577fedfe1de59c47
02:53:48:WU01:FS01:Uploading 2.33KiB to 171.67.108.102
02:53:48:WU01:FS01:Connecting to 171.67.108.102:8080
02:53:48:WU01:FS01:Upload complete
02:53:48:WU01:FS01:Server responded WORK_ACK (400)
02:53:48:WU01:FS01:Cleaning up
02:54:26:WU02:FS00:0xa7:Completed 19200 out of 80000 steps (24%)
02:55:26:WU02:FS00:0xa7:Completed 20000 out of 80000 steps (25%)
02:56:26:WU02:FS00:0xa7:Completed 20800 out of 80000 steps (26%)
02:57:26:WU02:FS00:0xa7:Completed 21600 out of 80000 steps (27%)

 

[iwajabitw]

If its a Nvidia card, drop down to the 373.06 drivers its a known issue with f@h.

Heres a link about it.

https://foldingforum.org/viewtopic.php?f=80&t=29276

 

[.far]

Thanks - am running 375.95, so will uninstall and get the older ones

 

[iwajabitw]

At least its an easy fix. Geforce Experience keeps bugging me to upgrade, but not until after the race. I had the same issue as you with my 980's

 

[.far]

What's the best way to downgrade the drivers? I tried uninstalling driver on one machine and it's now hung on uninstalling miracast driver, and the other one I tried the older driver doing a clean install while it thought it was happy - it is returning nothing to the screen after a few seconds of Win10 spinning dots. Well - apart from a heap of command line stuff the first time - now just nothing.. Hope I haven't borked these, they are both work based machines... D'Oh

 

[Markfw]

I did this to 10 win 10 computers. I just installed the new ones right over the old ones, and no problem.

Oh, but disable windows update first.

 

[TennesseeTony]

You can download DDU (display driver uninstaller) and completely wipe out any driver remnants if needed.

 

[.far]

Geez it was in a state - wouldn't even safe boot - just came up with half a screen of command line detail about watchdog initiation and starting reactor and froze.

But - luckily it's come good after swapping out the 2x1070s for an old AMD card - which allowed it to boot up and finish doing whatever was causing it's grief when the nvidia cards were in there, then swapped the cards back around and all fine! Silver lining on the cloud was that with the GPU's out, saw the PSU fan was dead so swapped that out too

Thanks for the advice/suggestions! No idea why my machines are so tetchy. The second one I'll let reboot in another day or so (it's used all the time..), and see what it does, but at least if it ends up in the same state I'll know how to solve it.