|
It's hard to say exactly what the problem is. I did find one thing of interest, however. I'm not sure if they're related, but it does indicate a potential problem.
Of your three SoB tasks, there's that failure and two others that completed and are waiting for wingmen. One looks normal, but the other includes this in the output prior to the actual result of the calculation:
Iter: 2006016/31304541, ERROR: ROUND OFF (0.5) > 0.4
Continuing from last save file.
Resuming Proth prime test of 24737*2^31304527+1 at bit 2003713 [6.40%]
Using all-complex FMA3 FFT length 2880K, Pass1=384, Pass2=7680, a = 3
That's not normal and indicates a calculation error. That's bad. LLR tried to recover the calculation by aborting and restarting from the last checkpoint file. It might or might not have been able to save the calculation by doing that. You'll know when the wingman returns a result to compare to yours.
Since this is a Skylake CPU, (which are known to have trouble with this type of computation), if you haven't already done so you should check with your motherboard manufacturer to see if there are any BIOS updates. The Skylake problem appears to be an issue with properly reading or using the memory configuration properly, and is fixable via the BIOS.
It's also possible that the calculation error comes from some other cause such as excessive overclocking or overheating, but since this is a known problem with Skylakes I'd certainly make sure the BIOS is up to date.
It's not clear to me whether this is also responsible for the error that your saw on the failed SoB task.
____________
My lucky number is 75898524288+1 | 
