Project: 5900 (Run 0, Clone 632, Gen 67) - Self-test failure

[toTOW]

Another Self test failure that my GTX+ does so well :

Code: Select all

[23:04:22] Project: 5900 (Run 0, Clone 632, Gen 67)
[23:04:22] 
[23:04:22] Assembly optimizations on if available.
[23:04:22] Entering M.D.
[23:04:28] Tpr hash work/wudata_08.tpr:  3392205633 1275481647 1148565387 620086690 602689319
[23:04:28] Working on Protein
[23:04:29] mdrun_gpu returned 
[23:04:29] Self-test failure
[23:04:29] 
[23:04:29] Folding@home Core Shutdown: UNSTABLE_MACHINE
[23:04:32] CoreStatus = 7A (122)
[23:04:32] Sending work to server
[23:04:32] Project: 5900 (Run 0, Clone 632, Gen 67)
[23:04:32] - Read packet limit of 540015616... Set to 524286976.
[23:04:32] - Error: Could not get length of results file work/wuresults_08.dat
[23:04:32] - Error: Could not read unit 08 file. Removing from queue.

There's no data in the DB yet.

[ihaque]

I forget - this was a card for which MemtestG80 returned no errors?

[toTOW]

No, this card never returned errors in MemtestG80. But sometimes, it fails to start a WU with this error, whatever core it uses (but core 11 1.19 has much greater chances to fail with this error), and I don't know why

[toTOW]

Someone else has been able to complete this WU successfully ...