Project: 10501 (Run 165, Clone 1, Gen 884)

[Napoleon]

Is this actually a bad WU? The error comes from an ancient machine with an old 9400GT card. But the GPU OC is pretty mild, shaders from 1375MHz stock to 1500MHz. GPU core is actually slightly underclocked, from 550MHz to 500MHz. GPU memory is at stock. Temperature should be pretty much OK, about 63C. CPU is at stock.

Code: Select all

09:37:00:WU01:FS00:Connecting to assign-GPU.stanford.edu:80
09:37:01:WU01:FS00:News: Welcome to Folding@Home
09:37:01:WU01:FS00:Assigned to work server 171.67.108.21
09:37:01:WU01:FS00:Requesting new work unit for slot 00: RUNNING gpu:1:"G96 [GeForce 9400 GT]" from 171.67.108.21
09:37:01:WU01:FS00:Connecting to 171.67.108.21:8080
09:37:02:WU01:FS00:Downloading 61.88KiB
09:37:03:WU01:FS00:Download complete
09:37:03:WU01:FS00:Received Unit: id:01 state:DOWNLOAD error:NO_ERROR project:10501 run:165 clone:1 gen:884 core:0x11 unit:0x000007366652eda54b6ea7ad00004074
09:37:13:WU01:FS00:Starting
09:37:13:WU01:FS00:Running FahCore: "C:\Documents and Settings\JML\Omat tiedostot\FAH\Client/FAHCoreWrapper.exe" "C:/Documents and Settings/JML/Omat tiedostot/FAH/Data/cores/www.stanford.edu/~pande/Win32/x86/NVIDIA/G80/Core_11.fah/FahCore_11.exe" -dir 01 -suffix 01 -version 702 -lifeline 160 -checkpoint 30 -gpu 0 -verbose
09:37:13:WU01:FS00:Started FahCore on PID 2016
09:37:14:WU01:FS00:Core PID:3784
09:37:14:WU01:FS00:FahCore 0x11 started
09:37:15:WU01:FS00:0x11:
09:37:15:WU01:FS00:0x11:*------------------------------*
09:37:15:WU01:FS00:0x11:Folding@Home GPU Core
09:37:15:WU01:FS00:0x11:Version 1.31 (Tue Sep 15 10:57:42 PDT 2009)
09:37:15:WU01:FS00:0x11:
09:37:15:WU01:FS00:0x11:Compiler  : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 
09:37:15:WU01:FS00:0x11:Build host: amoeba
09:37:15:WU01:FS00:0x11:Board Type: Nvidia
09:37:15:WU01:FS00:0x11:Core      : 
09:37:15:WU01:FS00:0x11:Preparing to commence simulation
09:37:15:WU01:FS00:0x11:- Looking at optimizations...
09:37:15:WU01:FS00:0x11:DeleteFrameFiles: successfully deleted file=01/wudata_01.ckp
09:37:15:WU01:FS00:0x11:- Created dyn
09:37:15:WU01:FS00:0x11:- Files status OK
09:37:15:WU01:FS00:0x11:- Expanded 62849 -> 336763 (decompressed 535.8 percent)
09:37:15:WU01:FS00:0x11:Called DecompressByteArray: compressed_data_size=62849 data_size=336763, decompressed_data_size=336763 diff=0
09:37:15:WU01:FS00:0x11:- Digital signature verified
09:37:15:WU01:FS00:0x11:
09:37:15:WU01:FS00:0x11:Project: 10501 (Run 165, Clone 1, Gen 884)
09:37:15:WU01:FS00:0x11:
09:37:15:WU01:FS00:0x11:Assembly optimizations on if available.
09:37:15:WU01:FS00:0x11:Entering M.D.
09:37:21:WU01:FS00:0x11:Tpr hash 01/wudata_01.tpr:  1678397534 2965368170 2433647248 1921020660 1365057275
09:37:21:WU01:FS00:0x11:
09:37:21:WU01:FS00:0x11:Calling fah_main args: 14 usage=100
09:37:21:WU01:FS00:0x11:
09:37:23:WU01:FS00:0x11:Working on Protein
09:37:23:WU01:FS00:0x11:Client config unavailable.
09:37:24:WU01:FS00:0x11:Starting GUI Server
09:47:12:WU01:FS00:0x11:Completed 1%
09:57:02:WU01:FS00:0x11:Completed 2%
10:06:51:WU01:FS00:0x11:Completed 3%
10:16:40:WU01:FS00:0x11:Completed 4%
10:26:29:WU01:FS00:0x11:Completed 5%
10:36:19:WU01:FS00:0x11:Completed 6%
10:46:08:WU01:FS00:0x11:Completed 7%
10:55:57:WU01:FS00:0x11:Completed 8%
11:05:47:WU01:FS00:0x11:Completed 9%
11:15:37:WU01:FS00:0x11:Completed 10%
11:25:26:WU01:FS00:0x11:Completed 11%
11:35:15:WU01:FS00:0x11:Completed 12%
11:45:05:WU01:FS00:0x11:Completed 13%
11:54:55:WU01:FS00:0x11:Completed 14%
12:04:44:WU01:FS00:0x11:Completed 15%
12:14:33:WU01:FS00:0x11:Completed 16%
12:24:23:WU01:FS00:0x11:Completed 17%
12:34:12:WU01:FS00:0x11:Completed 18%
12:44:01:WU01:FS00:0x11:Completed 19%
12:53:51:WU01:FS00:0x11:Completed 20%
13:03:40:WU01:FS00:0x11:Completed 21%
13:05:04:WU01:FS00:0x11:Run: exception thrown during GuardedRun
13:05:04:WU01:FS00:0x11:Run: exception thrown in GuardedRun -- Gromacs cannot continue further.
13:05:04:WU01:FS00:0x11:Going to send back what have done -- stepsTotalG=15000000
13:05:04:WU01:FS00:0x11:Work fraction=0.2114 steps=15000000.
13:05:08:WU01:FS00:0x11:logfile size=14427 infoLength=14427 edr=0 trr=23
13:05:08:WU01:FS00:0x11:+ Opened results file
13:05:08:WU01:FS00:0x11:- Writing 14963 bytes of core data to disk...
13:05:08:WU01:FS00:0x11:Done: 14451 -> 4497 (compressed to 31.1 percent)
13:05:08:WU01:FS00:0x11:  ... Done.
13:05:08:WU01:FS00:0x11:DeleteFrameFiles: successfully deleted file=01/wudata_01.ckp
13:05:08:WU01:FS00:0x11:
13:05:08:WU01:FS00:0x11:Folding@home Core Shutdown: EARLY_UNIT_END
13:05:09:WARNING:WU01:FS00:FahCore returned: BAD_WORK_UNIT (114 = 0x72)
13:05:09:WU01:FS00:Sending unit results: id:01 state:SEND error:FAULTY project:10501 run:165 clone:1 gen:884 core:0x11 unit:0x000007366652eda54b6ea7ad00004074
13:05:09:WU01:FS00:Uploading 4.89KiB to 171.67.108.21
13:05:09:WU01:FS00:Connecting to 171.67.108.21:8080
13:05:10:WU01:FS00:Upload complete
13:05:10:WU01:FS00:Server responded WORK_ACK (400)
13:05:10:WU01:FS00:Cleaning up

[bollix47]

Another folder was able to finish it for full credit:

Hi xxxxxxxxxx (team xxxxxx),
Your WU (P10501 R165 C1 G884) was added to the stats database on 2012-12-26 02:04:54 for 587 points of credit.

[Napoleon]

Thanks for checking up on the WU - so much for even the slightest overclock, then. I did stability testing before folding, of course, and the GPU has folded many WUs just fine before this one. But with such an old card, I'm not even going to 2nd guess the reason for this failure. Back to stock clocks all the way, and if there's even one more "Another folder was able to finish..." thingy for this card, I'll declare it as junk.

[bruce]

@Napoleon:
Had you been running any other GPU application?

I've recently become aware of potential problems running out of VRAM when some other application allocates too much memory and then doesn't clean up after itself. The error messages don't explain WHY the WU failed, only that it did, and they're not particularly helpful when strange things like that happen. GPU sharing is still an emerging technology.

[Napoleon]

Totally dedicated to FAH as soon as I got it, but this is a 2nd hand setup a friend donated to me for giving some advice when he decided on buying a new 'puter. The GPU in question actually has 512MB VRAM, which I consider sufficient.

As far as VRAM goes, I've tried reducing VRAM to 256MB in mobo BIOS of my ION GPU. If let be, folding is just fine. But as it is subject to several mid-WU pause/restart cycles, I noticed intermittent failures I simply don't understand. Recovery by reboot required. My solution: increase VRAM back to 512MB, problem solved.

Anyway, back to this particular problem: it was a system which had been going on with major dust-build problems and a marginal PSU. Only casually used, but still... this doggy won't last long, it's just a matter of fond farewell and a graceful death after many years of faithful service.

*Gollum note to self: stop pampering abused hardware. Rings suck big time.