I have a few rigs with 9GPU AMD 4GB and all of them have the same problem. Only 8 GPU's are working, GPU 1 hashrate is 0.0MHS and after some time miner restarts because GPU "is not working". Any tips for that error? Tried latest windows, different drivers....etc If I try to mine different coin they are all wrking at full speed, and no issue with GPU1....
You can also use integrated GPU if your CPU/Mobo supports it. Just connect hdmi to motherboard and check it out. Windows will use iGPU to display desktop and Phoenix will be able to use full GPU memory.
I had almost the same problem (my last card wasn't mining full hash instead of your first) and I was crying like a baby here
ONLY solution I found was to update driver to 20.7.1 or 20.8.1
Problem with those drivers is Claymore won't work if you use Claymore for injecting straps...well it's not a problem, it's just something I can't use any more to get max hash rate from my 4 GB cards.
Also, another problem even when I solved this issue was, Team Red Miner not recognizing compute mode turned ON and patching registry...
After that, I was back to "last card not working" issue.
So, update to last driver, no more Claymore for injecting straps and no TRM because it just messes up everything. (for me)
I tried all drivers from list Phoenix gave (18.12.1.1 to 20.8.1), whole week without normal sleep, trying everything
If you know how to use AMD Mem Tweak you can inject straps from there...I was getting whole 10Mh/s from Claymore on my 6x rx470s rig, so it's not a small gain.
Also, try not to mine on Nicehash with 4GB GPUs because they very often switch from ETH to ETC...switching from ETH to ETC and back to ETH makes you loose a tons of time for Phoenix to get back to normal hashrate - really annoying
Phoenix sometimes can skip whole ETC DAG Nicehash mess, with "not enough memory for DAG" and restart miner back to ETH...but that's if you're lucky and your Phoenix goes on lucky "miner restart" streak
Everything I wrote is something I experienced, not saying it will happen to you.
At the same time, second rig with exact same 6x rx470s from the same batch, same custom vbios, is working like a charm on 18.12.1.1 driver, so I guess my Gigabyte H110 mobo is the source of all my problems - only thing I couldn't check because it's a long drive, so I only suspect it
Hope you'll get all your cards back to full work mode only with driver update