Amateur Miner from sunny South Africa here, and I need your help.
I’ve recently made the switch from Windows to HiveOS, I have 3 workers and all was running well for a few days. Then one day after a reboot, one of my RX580’s just stopped working.
On every reboot since, I get errors: ERROR ring gfx test failed (-110) ERROR hw_init of IP block <gfx_v8_0> failed -110
amdgpu_device_ip_init failed
Fatal error during GPU init
Then in Hive interface I see the card, but no temp/fan/power values are displayed, and then claymore doesn’t even show that card (only GPU 0-4, i.e. the first 5 cards)
Some background:
Total of 3 rigs, this rig has 4x RX580 and 2x RX570 (all mixed brands and memories)
This failing card is an ASUS Dual OC 8GB with Hynix memory and was running with pimped straps and mild OC/UV (1150/900/2100) and was comfortable at ~28MH/s.
After failure I attempted the following:
Removed OC (no change)
Flashed default BIOS (which I always save from card before modding) (also no change)
Swapped riser with working card (no change)
Swapped 8-pin power with working card (no change)
Pulled card and put into Windows 10 machine - booted into Windows with “Standard VGA Adapter” driver BUT when installing AMD drivers instant BSOD (tried multiple driver versions/each failure removed using DDU) (obvious failure)
Put card back into HiveOS rig, flashed a modded BIOS from tech forum, and got stats (temp/fan and power) to display, but no mining. Flashed original back on with no OC. Same result (stats but no mine). After a few reboots, it is back to the main issue described above.
Flashed this same modded BIOS (from the forum) to see if stats come back at least but to no avail. Still completely useless except the name shown in HiveOS interface.
Has anyone seen/experienced this before? Please help.
Well by the looks of things, it seems the card itself is failing. There’s not a whole lot that can be done aside from what you’ve tried already.
You reverted to default factory BIOS, removed OC, issue is seen on multiple OSes (so this excludes software/drivers + other cards are working as intended, so this should already be a good give away), you’ve swapped risers and cables (excludes other potential hardware)…
I’m having the same issue in my test rig. I have a 580 and 570 doing it. They work fine when I move them to another rig and changing the riser or which PSU they are plugged into didn’t change anything. I feel like it’s driver related with the OS. Going to mess with it more tomorrow.
Hello,
Fist try phoenix miner. If that doesn’t help, try switching to a different pcie port on your motherboard. I have had this issue before and it wasn’t a failing gpu.
Good luck (also with eskom’s kak)
I’m experiencing the exact same thing I even started a duplicate thread from what I can tell
This started happening after I repasted and repadded my card so leaning heavily towards it being broken.
I also tried different motherboards, different operating systems to no avail.
Sorry for the late reply. So it turns out that one or more chips are fried on the cards giving these errors. I sold mine to a more capable chap to solder replacement chips.
