03-25-2017, 12:06 AM
I know this is a OpenCL/hardware issue and not a hashcat, but thought you good folks might have some insight.
Quick specs:
Ubuntu 14.x LTS
NVIDIA driver v.375.39
Intel i7-6850, 32 GB RAM
3x EVGA 1070's
Gigabyte x99 gaming something mobo
1500W Corsair PSU.
*no* risers!
Temperatures stay cool/OK.
After a some time it will drop one of my GPUs
clinfo does show that its gone. But
Show all three.
A reboot (even a soft reboot, aka: reboot from a prompt) gets it back, for a while.
How long it lasts is as random as can be.
Usually when it hangs its, at the beginning during initializing memory (?, initializing something)
I can do a kill -9 and it doesn't exactly stop hashcat, but gets it running with one dropped GPU
All further runs (without a reboot) will run fine with just the 2 GPUs.
Any ideas?
Is there a way to 'jump-start' OpenCL to get it back without a reboot? (as a work around)
TIA
Quick specs:
Ubuntu 14.x LTS
NVIDIA driver v.375.39
Intel i7-6850, 32 GB RAM
3x EVGA 1070's
Gigabyte x99 gaming something mobo
1500W Corsair PSU.
*no* risers!
Temperatures stay cool/OK.
After a some time it will drop one of my GPUs
clinfo does show that its gone. But
Code:
lspci -vnn | grep VGA -A 12 | grep VGA
A reboot (even a soft reboot, aka: reboot from a prompt) gets it back, for a while.
How long it lasts is as random as can be.
Usually when it hangs its, at the beginning during initializing memory (?, initializing something)
I can do a kill -9 and it doesn't exactly stop hashcat, but gets it running with one dropped GPU
All further runs (without a reboot) will run fine with just the 2 GPUs.
Any ideas?
Is there a way to 'jump-start' OpenCL to get it back without a reboot? (as a work around)
TIA