10-04-2019, 06:05 AM
So I was running multiple tests with 4 GPUs. I got some good benchmarks and decided to give it room to stretch. A couple of test hashes and wordlists popped out without issues. But when I fed it a true dictionary with OneRuletoRuleThemAll, it spun for about five minutes then lost one of the GPUs. Another couple minutes later the fourth GPU was reporting 0% utilization. I killed all processes and halted the machine.
From posts online, I figured it might be USB3.0 cabling for the risers, so I swapped all cabling, and even swapped some GPUs and PCI slots.
At this point now, only two GPUs come up hashcat's identification & when trying to run anything I'm getting the "CL_PLATFORM_NOT_FOUND_KHR" error, which I believe is usually driver related.
Why is this happening now, when the rig was running (limping) along earlier?
& Still looking to see if anyone has a suggestion for my hard 4 GPU limit above...
From posts online, I figured it might be USB3.0 cabling for the risers, so I swapped all cabling, and even swapped some GPUs and PCI slots.
At this point now, only two GPUs come up hashcat's identification & when trying to run anything I'm getting the "CL_PLATFORM_NOT_FOUND_KHR" error, which I believe is usually driver related.
Why is this happening now, when the rig was running (limping) along earlier?
& Still looking to see if anyone has a suggestion for my hard 4 GPU limit above...