Folding Forum

.

osgorth wrote:

Code: Select all

[00:57:15] Folding@Home GPU Core - Beta
[00:57:15] Version 1.18 (Mon Oct 13 11:11:30 PDT 2008)
[00:57:15] 
[00:57:15] Compiler : Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 
[00:57:15] Build host: amoeba
[00:57:15] Board Type: AMD
[00:57:15] Core : 
[00:57:15] Preparing to commence simulation
[00:57:15] - Looking at optimizations...
[00:57:15] - Created dyn
[00:57:15] - Files status OK
[00:57:15] - Expanded 68504 -> 357580 (decompressed 521.9 percent)
[00:57:15] Called DecompressByteArray: compressed_data_size=68504 data_size=357580, decompressed_data_size=357580 diff=0
[00:57:15] - Digital signature verified
[00:57:15] 
[00:57:15] Project: 5747 (Run 1, Clone 11, Gen 17)
[00:57:15] 
[00:57:15] Assembly optimizations on if available.
[00:57:15] Entering M.D.
[00:57:21] Working on Protein
[00:57:21] Client config found, loading data.
[00:57:21] Starting GUI Server
[00:57:24] mdrun_gpu returned 
[00:57:24] Nonzero force sum on GPU
[00:57:24] 
[00:57:24] Folding@home Core Shutdown: UNSTABLE_MACHINE
[00:57:27] CoreStatus = 7A (122)

I am getting similar results with the 47XX series WU's with my HD3870 and catalyst 8.11. 620r1 MPICH console client. Oddly the 57XX series is the only one that so far will progress passed the Starting GUI. The 1.18 was terminating in EUE waiting 24 hrs ( all other tha 57XX) before I updated to 8.11 driver. I have only been using this card for a little over a day as just installed. Since 8.11, F@H 1.18 tries 3 times with a 47XX and then I have been getting a 57XX before EUE termination.

Completed successfully too :

Hi osgorth (team 37451),
Your WU (P5747 R1 C11 G17) was added to the stats database on 2008-11-22 20:45:36 for 384 points of credit.

You're correct, it was later picked up and successfully folded!?
This is new behaviour - first time it crashes instantly, then a few hours later it picks up the same WU and folds it without a problem. Isn't that strange??