yeah i'll fix it, but i'm at work and di this by remote. my mistake for doing this by remote. bios is not modded since its not possible.
second bios is only ULPS on these cards wich reduces hashrate alot so no. It wont even start the cards thats found. so it's sutdown for now
Ok
so i only suspect the pool or a drop of hashrate on the miner. Try with Cast_xmr or xmr-stak.
Check if you have the last 18.3.4 Adrenaline drivers, compute mode enabled. and for Cast_xmr use --algo=1 --ratewatchdog --forcecompute
well, i'm still using the 17.30.1029 beta blockchain driver. although i did not have this problem on xmr with xmr-stak. Is the 18.3.4 driver better?
I havent more little drop hashrate, same results than the blockchain drivers but much more stability with high overclocking or multitasking.
On XMR-stak you can try to :
Down a little the intensity to see if the same.
Configure 2 threads each card (more stable and hashrate)
Set affine_to_cpu : 1 if you are mining with CPU at the same time (more stable)
With my vega i have 2055H/s each card with Cast-xmr.
I'll try the new driver when i ot time to install it,
the rest of your suggestions is already done and tweaked.
intensity is lower for 2 cards that got obvious errors, the rest is running like a charm at 1932.
2 threads per cards is basic knowledge for anyone with vegas
affine_to_cpu 1 is set, and core 1 is removed from cpu.txt
i'm not even pushing the cards that hard, 920 MHz for 56 and 1000 for 64.
gpu mem temp at 55 C
Like i said, i had none of these problems mining xmr, only stellite communitypool thats giving me 25% less than i should get. sure, could be that the driver has been acting up since its very flaky. For now looks like i'm switching to xmr in order to evaluate. 25% is a lot!
i might try a different pool just see if its the pool or not...