Before I took your advice, my rig lost connection and cards stopped reporting. I tried to update Nvidia drivers remotely and may have messed up my 1660S. I just replaced the flash with a new one/clean version of HiveOS. Besides the Nvidia card, everything is fine after following your guidance. I haven’t flashed the ubermix 3.2 yet. Would you recommend reverting all Samsung’s to their original bios and then flashing the new bios or can I just go ahead and flash the modded bios to them now? (After stopping the miner obviously)
Also, do you have a public ETH wallet address that is currently active on a pool? I’d like to mine to your wallet for a day or two for you help.
1 Like
0x9ff1840a9a93ca7dcafc78ce703374f8454c7f94
This is the wallet. Any help appreciated.
There is no need to flash original back. You can use your modded and run it again through polaris. It will replace uber mix 3.1 with uber mix 3.2 if you follow correctly the steps. No to uber 3.1 yes to 3.2.
Hey I noticed that Memory Controller Voltage is apparently reserved only for “Navi” cards and mine are Polaris. Should I ignore that? Also, I’m having temp/load (I’m assuming the percentage adjacent to the temp is the load?) issues with two cards if you could check it out. GPU 1 has ubermix, GPU 2 didn’t provide the option in PE so it doesn’t.
Dude memory controller voltage is working also for polaris. Only memory voltage cant be changed in polaris. I see you have a problem with two of the cards. 80 is too high temp. They are for cleaning and change of thermo paste or thermo pads
Your cards are doing too high watts as if they dont overclock correctly. Open console and write amd-info and post the results. They need to be below 100 watts all.
Make the fans 100% until we resolve the issue
1 Like
Fans @ 100% and core temps have stabalized. Hashrates fell for GPU 1 & 2, they probably need thermal paste. The amd-info was taken before applying 100% fans. Rig is currently mining to your address btw
Restart the rig. It appears they dont work on the set core voltage but higher. Restart and do amd-info again
Is GPU 3 using Core State 3? I don’t have Core States set, as per your instructions. But in amd-info it seems to be saying that it’s using Core State 3, which is the number in bold with an asterisk (*), right?
They all are using some state as indication. It will show in amd info. But still you have a problem with the OC. They just dont get it.
Upgrade hive to latest version. And we will see again how they perform.
Thanks for mining to my address. I appreciate it.
Let me know how is it going. I will be here again after 7 hours.
Hey everything’s running well after update and reboot. Fans still @ 100%. I may still apply thermal paste to some of the cards. We’ll see if they spit out any invalids over the next 12hrs.
Post a pic of the OCs to see the watts temp and values
Still high watts. I will read in internet how to fix this. They need to be under 100. Temps will fall when we do this.
I think you need to reinstall the hiveOS. flash it again and configure it again. Some other dude wrote in the forum it doesnt take OCs and after flashing fresh the hiveOS it fixed the problem for him.
1 Like
alright cool, I’ll do it tomorrow. Still want to reapply thermal paste too