I have (4) BTC-S37 (8 GPU) mining boards full of RX5700XT. I can get them to mine on all 8 cards for a period of time. Then occasionally, they will crash and lock up. If I don’t catch it in time and they are off for too long. When I finally go to start them up HiveOS freezes up as soon as it gets to the AMD drivers of the boot. It will show AMD drivers 8 GPU and then freeze. Once this happens I can not get HiveOS to come online again or the miner to load. I have to pull all the cards, clear the CMOS and flash a clean version of HiveOS on the SSD. Then it is a struggle to get it to mine again. I have to load the board 1 card at a time, until i finally get all 8 cads to mine again. I’m desperate for some guidance here as I’m at my wits end and ready to pull my hair out!! Any relevant information would be greatly appreciated!
Try to set your mining so it doesn’t start for 2 minutes after boot.
If you can get that far I have found if my “hashrate monitoring” is too quick it sometimes locks up. I have 4 RX 5700 XT + 1 2060 on a BTC-S37 no risers with 1 PSU and I found that give me more stability. I have no monitor attached to it and I have a smartplug that I can turn on/off from my phone so if it locks up I can reboot it. Memory? I had 4GB but increased to 8GB as over time after the 3rd of 4th day of uptime I noticed the free RAM was heading towards 0 faster than I would have thought.