So this issue has been plaguing me for the past week. I am having an issue with at Random Times, my mining rig will just completely Freeze out of no where. And by freeze I mean the mouse stops moving on the screen when I move the mouse, and everything is just frozen there. Rig stays powered on, can’t be pinged, looses connecting to hive servers and is totally unresponsive. Sometimes it happens after an hour of farming, other times it happens after 8 hours, other times it happens after 18 hours, its random, which makes this hard to trace. I’m going to list my current setup below, but first, I will say this… I started by changing out the RAM, thinking that was the issue, but it didn’t help. Then I changed the hard drive, and rebuilt the image, still same issue. In my final plea for a fix, I ended up changing EVERYTHING!! Motherboard, Ram, Hard Drive, and even power supply.
Ok so here is my setup.
I have 6 NVIDIA RTX 3060 GPU’s They are not all the same, some are EVGA, one ASUS, one PNY, etc… I plan on buying two more cards to make 8 total cards in this Rig.
Motherboard: ASUS Prime Z590-V
Processor: Intel Pentium Gold G-6400
Ram: Corsair Vengeance LPX 16GB (2x8GB) DDR4
Hard Drive: Samsung NVME 250GB Drive
Power Supply 2 X EVGA 850 B5, 80 Plus BRONZE 850W
I made sure to keep the pair up the Card / Riser connected to the same power supply, I was told connecting the riser to one PS, and the Card to another was a bad idea. All Risers are powered with a 6 pin PCIE power connector, NO SATA power!!!
Riser Cards: I’m using two Pcie Riser Cards 1 x 4, I am not sure of the brand, as its not stated on the card. I have 4 GPU’s connected to one card, and 2 GPU’s connected to the second card. Not sure if this is allowed, but here is the link to the 1 x 4 card and Risers I am using.
Riser Cards on Amazon
I had originally started with NONE of the cards overclocked, but then as an attempt to try and figure out what was happening I thought of trying to lower the temp / power usage in case that had something to do with it. So my current OC settings are as follows. (I am no overclocking Pro)
Fan: 75
Core: 1600
Memory: -300 (I did this thinking it would keep the cards cooler, at this point I didn’t care about hash rate, just wanted to cool down the cards thinking that may be the issue, Is this bad?)
Power: 130
I am using the Newest version of HiveOS. Ive tried upgrading to a newer version every time one came out thinking “Maybe this is a bugged version” but after upgrading through the last 5 versions, I still have the same issue.
I am using NBMiner with no additional parameters, and I am mining ETH to the Ethermine Pool
When the miner does run, with all the settings above, I get about 30 MH/s Core temps on the card are ranging between 40 to 50 degrees currently. Cards are ranging from between 95 to 105 watts of power usage.
Now obviously before I did any of the OC changes, the card were running between 55 to 62 degrees and using about 130 watts of power. I did the OC changes to see if reducing the temp / power usage would help.
As far as power goes. I have two 850 Watt Power Supplies, as listed above, One of them is powering 3 cards / risers only, The second power supply is powering 3 cards / risers and the motherboard. I have a Watt Meter connected to the rig, and with all 6 cards running, I am at about 900 watts of power used. which is well below the rating of the two power supplies. From my math it doesn’t seem like 3 3060 Cards and risers would take anywhere near 850 wats of power, so I don’t believe its an issue with not enough power.
I hope I didn’t miss providing any needed information.
I am asking for any help in trying to figure out what might be the issue here. I am even willing to pay someone to do some one on one time with me to get this figured out if I need to.
Any info / help is appreciated. Thank you in advance.