Yeah-I saw that about driver-I was getting best performance and stability on 495.44. I reverted this a 470 version for some testing and when I went to replace the 495.44, it also was not on the list. There were a couple OS updates this week, maybe some users are having issues with it, but it’s good for me.
Fortunately I had downloaded the driver package previously so was already on my system (but not on HiveOS available drivers list). nvidia-driver-update —list will bring up available, but if you had 495.44 installed before, the package might still be on your system-it says it at the very bottom-if you have the driver, it will update if you select it.
İ have little problem guys someone’s can help me
Gpu 1 hung detected! error - if this error happened several times, or failed to detect some gpus, please check your hardware.
Gpu 1 its 3060ti LHR how can i fix this.
Is it on a riser or plugged in directly? Recheck connections straight and good.
Are your bios settings good (gen 2)? Does the PCIE port you are using support your card?
Do you have power applied to both available power input ports of the GPU?
From command prompt at terminal, nvidia-info command should see your card and know about it…, I’d force a driver reinstall either way (nvidia-driver-update —force).
Post back if you are still having issues, but those are what I’d be looking at.
I did not do it yet, but I saw several opinions that say that it is the solution, for hiveos it is very easy although I do not encourage me yet, I already have the bios downloaded, I know that it lowers the consumption and improves the mhs
Ok-also, I didn’t fully appreciate the temperature/performance correlation, it matters.
So I had my rig running very reliably at around 39 MHS, most of the time, so I wanted to park it in the rafter area of my garage-less airflow and higher ambient temps.
Everything was mostly good at night and early AM when ambient temps are the lowest, but in the afternoons, like 2-6 PM, my hash rates would dip down to the 235 MHS range as the card temps increased.
So of course, these things matter over time, so I busted the rig back down to ground level and lower temps. Up until now I’d been using auto fan settings, but I was like, what if I disable auto fans, and run the card fans at 60%. Sure as shit, I picked up a few MHS.
Yes, check the same, at this moment I have the Pl in 135 and I lost some more, I use another oc that seems better to me, anyway the definite solution has to be the change of bios
Hey my man, how’s your rig going? I have 8 of these guys…. Exact same ones you have…. But I can’t get any to even touch 40……(except 1… barely) any thoughts or input ya could pass on would be greatly appreciated!
Current Nvidia driver (495.44 and Hive OS versions)
Core -250
Mem 1800
PL 140
NBMiner-40.1
Give it a shot and keep them cool.
One or 2 of mine stay over 40, other four are between 38 and 40.
The cards fully unlocked should do about 50 MHS, my initial values on a different non-LHR optimized miner (Phoenix) we’re about 25 MHS, so if you are getting anything north of 37.5 MHS, you are seeing 75% full capacity.
I think that is the best you can expect without exotic and risky hacks. I’m exploring those cautiously, hoping mining software can get better LHR performance.
I have (4) 1660 series cards I was going to sell, hoping these (6) 3060 cards would give me around 300 MHS, but with LHR issue, I’m gonna invest in a new 8 rig MoBo/Power supply, put my (6) 3060’s into that and put my (4) 1660s back online. I’ll have 4 slots available for additional GPUs, and about 350 MHS total output.
Just managed to get 48+ Mhs on my Asus RTX 3060 Ti
But again one of the cards are really tricky to get stable above 44 Mhs
The one on 44 was on just 38 and hitting the LHR.
I flashed it with the vBIOS from the other one and now its alot higher and not hitting the LHR limiter
