Page 1 of 1

Nothing but Bad WUs for my GPU

Posted: Sun Jun 30, 2013 9:48 pm
by LonePalm
I followed the instructions for Bad work units above. No fix.

I have had over 2000 bad work units today and despite successful turn ins, no points.

Here is my log file.

Code: Select all

# Windows GPU Console Edition #################################################
###############################################################################

                       Folding@Home Client Version 6.41r2

                          http://folding.stanford.edu

###############################################################################
###############################################################################

Launch directory: C:\Folding @ Home\GPU Console
Executable: C:\Folding @ Home\GPU Console\Folding@home-Win32-GPU.exe
Arguments: -verbosity 9 

[21:42:10] - Ask before connecting: No
[21:42:10] - User name: LonePalm (Team 36120)
[21:42:10] - User ID: 4E6C374F748E1D8F
[21:42:10] - Machine ID: 2
[21:42:10] 
[21:42:10] Gpu type=3 species=20.
[21:42:10] Loaded queue successfully.
[21:42:10] - Preparing to get new work unit...
[21:42:10] - Autosending finished units... [June 30 21:42:10 UTC]
[21:42:10] Cleaning up work directory
[21:42:10] + Attempting to get work packet
[21:42:10] - Will indicate memory of 4087 MB
[21:42:10] Gpu type=3 species=20.
[21:42:10] - Detect CPU.[21:42:10] Trying to send all finished work units
 Vendor: GenuineIntel, Family: 6, Model: 14, Stepping: 5
[21:42:10] + No unsent completed units remaining.
[21:42:10] - Connecting to assignment server
[21:42:10] Connecting to http://assign-GPU.stanford.edu:8080/
[21:42:10] - Autosend completed
[21:42:11] Posted data.
[21:42:11] Initial: 40AB; - Successful: assigned to (171.64.65.69).
[21:42:11] + News From Folding@Home: Welcome to Folding@Home
[21:42:11] Loaded queue successfully.
[21:42:11] Gpu type=3 species=20.
[21:42:11] Empty passkey
[21:42:11] Connecting to http://171.64.65.69:8080/
[21:42:12] Posted data.
[21:42:12] Initial: 0000; - Receiving payload (expected size: 4373599)
[21:42:15] - Downloaded at ~1423 kB/s
[21:42:15] - Averaged speed for that direction ~1424 kB/s
[21:42:15] + Received work.
[21:42:15] + Closed connections
[21:42:15] 
[21:42:15] + Processing work unit
[21:42:15] Core required: FahCore_17.exe
[21:42:15] Core found.
[21:42:15] Working on queue slot 02 [June 30 21:42:15 UTC]
[21:42:15] + Working ...
[21:42:15] - Calling '.\FahCore_17.exe -dir work/ -suffix 02 -nice 19 -priority 96 -checkpoint 15 -verbose -lifeline 9700 -version 641'

[21:42:15] *********************** Log Started 2013-06-30T21:42:15Z ***********************
[21:42:15] Project: 8900 (Run 355, Clone 0, Gen 43)
[21:42:15] Unit: 0x00000043028c126651a6758ae977f8e3
[21:42:15] CPU: 0x00000000000000000000000000000000
[21:42:15] Machine: 2
[21:42:15] Reading tar file state.xml
[21:42:16] Reading tar file system.xml
[21:42:17] Reading tar file integrator.xml
[21:42:17] Reading tar file core.xml
[21:42:17] Digital signatures verified
[21:42:17] ERROR:exception: neither options[gpu] nor BRT_ADAPTER is set
[21:42:17] Saving result file logfile_02.txt
[21:42:17] Saving result file log.txt
[21:42:17] Folding@home Core Shutdown: BAD_WORK_UNIT
[21:42:22] CoreStatus = 72 (114)
[21:42:22] Sending work to server
[21:42:22] Project: 8900 (Run 355, Clone 0, Gen 43)
[21:42:22] - Read packet limit of 540015616... Set to 524286976.


[21:42:22] + Attempting to send results [June 30 21:42:22 UTC]
[21:42:22] - Reading file work/wuresults_02.dat from core
[21:42:22]   (Read 911 bytes from disk)
[21:42:22] Gpu type=3 species=20.
[21:42:22] Connecting to http://171.64.65.69:8080/
[21:42:22] Posted data.
[21:42:22] Initial: 0000; Conversation time very short, giving reduced weight in bandwidth avg
[21:42:22] - Uploaded at ~3 kB/s
[21:42:22] - Averaged speed for that direction ~2 kB/s
[21:42:22] + Results successfully sent
[21:42:22] Thank you for your contribution to Folding@Home.
[21:42:24] ***** Got a SIGTERM signal (2)
[21:42:24] Killing all core threads

Folding@Home Client Shutdown.

Re: Nothing but Bad WUs for my GPU

Posted: Sun Jun 30, 2013 10:35 pm
by bruce
First of all, I would recommend upgrading to the Version 7 client. It will recognize your GPU correctly and assign WUs that are suitable for it. Version 6 has not been tested with FahCore_17 and it's not surprising that you're having troubles. The WUs are not bad ... they're just not the proper assignment for whatever GPU you actually have.

You do not get points for WUs that are not completed.

Please stop the client immediately. Allowing it to dump 2000 WUs that are probably NOT bad is a bad idea.

Re: Nothing but Bad WUs for my GPU

Posted: Sun Jun 30, 2013 10:59 pm
by LonePalm
I have stopped the client. It dumped 2000 WUs before I knew anything was wrong.

Re: Nothing but Bad WUs for my GPU

Posted: Sun Jun 30, 2013 11:41 pm
by Jesse_V
How could it possibly dump that many WUs? Even the v6 client pauses itself for 24 hours to wait for the user to resolve the problem. The V7 client will disable the slot after 10 failed WUs. 2000 failed WUs in one day seems a stretch to me...

Re: Nothing but Bad WUs for my GPU

Posted: Sun Jun 30, 2013 11:43 pm
by LonePalm
Installed and running. How do I use my FAHMon to monitor my WU progress?

Re: Nothing but Bad WUs for my GPU

Posted: Mon Jul 01, 2013 7:17 am
by Napoleon
I think you'll have to switch to HFM for 3rd party monitoring. Client v7 doesn't support log file parsing anymore, and AFAIK HFM is the only 3rd party monitoring utility which has been upgraded.