Hello Guys . I use my 3060`s lhr1 (Ethash) . Have you idea how can i change oc settings or miner additional settings to have less rejected shares? I want keep the current hashrate ( or get ~ 37.5 - if its possible) and energy consumption.
Before that, I had set up Mem: 2300 and later 2000 but there were even more rejected shares but hashrate was 37.5,h/s.
Acá van mis números, es lo mejor que logre hasta el momento (y estable)
These are my numbers, it’s the best I could achieve so far (and stable)
3060 LHR V2, Zotac, 2 fan
Have a GeForce RTX 3060 12053 MB (2487 GA104) V2, also called a zotac 3060 twin edge. Tried several miners, and the best and most stable has been lolminer, using –mode LHRLP.
As for overclocking, the card has samsung memory and doesn’t like to be oc too much, otherwise, it freezes and crashes the computer. 2100 is as high as it will go (was tying to get the fan glitch), and probably 1800 for long-term stability.
The core doesn’t like relative values, as the power skyrockets. This requires a locked value. If too low the card never goes beyond the twenties. Around 1400-1500, I start to see 34+ MH/s.
Did try a full x16 gen 3.0 pcie slot, but made no difference. Can’t run the fabled 460.39, or the card crashes, so using the latest 470.74. Since I’m on hiveos 470.05 isn’t an option either.
Currently using 1500 core, 2100 memory, 115 PL, with 34.54 MH/s eating 100–110W (though the power oscillates a bit). Even though the hashrate is rather stable, hivepool reports it all over the place. You can save a little power by using 1400 core and 1800 memory.
You are getting 1 share a minute, I’m getting 3-4 shares a minute… same cards, but I have less aggressive oc and less power consumption… I use nbminer… core 1500, mem 2350, pl 120, fan 70 … max 105 W per card… average 33.50 mh per card
Do you mine on the pool closest to you? High ping (50+) can result in faulty shares
Do you have a wired and stable internet connection? Same as above, if you have a bad connection then try to get a better provider.
If the two above is yes, then:
Use lolminer 1.34 or NBminer 39.6
For LHR 3060’s, test your core clock between the 1400-1500 range and Samsung memory clocks between 2200 and 2700.
I have three 3060 rigs, one v1 LHR and two v2 LHR and I tried to go for stability so i’m using 1450 core 2200 memory and 103 PL on all of the cards and i manage to get 39 MH on the v1 cards, 34 MH on the v2 cards with an overall efficiency of 98.81% with lolminer 1.34
BTW! I saw that you are CPU mining Monero while doing GPU mining, your LA is not too high because of that? It might cause share drops If you overload the CPU.
If you use -lhrlp you have to set the core clock to 1150 or lower for the mode to take effect. Since your core OC is 1500 it’s pointless to leave the command running.
Samsung RAMs sure can take a beating, you can achieve clocks from 2200 up to 2700. I have alot of 3060 and all of them are stable on 2200, some can run on 2400 and some on 2600 but they crashed hard on 2800. Since it didn’t change much, I choose stability over +1 MH/s.
ATM I’m using 1450 core, 2200 memory and 103 PL on all of my cards, with this i can achieve 39 MH/s on the v1 LHR and 34 MH/s on the v2 LHR 3060’s and the power draw is between 88W and 103W
Every mining pool fluctuates your hashrate alot. If set correctly, your GPU’s hashrate is almost static (+/- a few decimals) in HiveOS. What you can really base on, is the 24h average of the pool.
For me, i have a 648 MH/s in HiveOS and the pool is always changing between 430 MH/s and 810 MH/s, while the 24h average is 642 MH/s, wich is acceptable.
Thanks! i am on lolminer 1.34ver and right now i am getting 39mh…
i got the nonLHR cards, maybe that helps…i dont know.
BUT like i complained many times, lolminer is very unstable! I am having constant problems and resets, and bad shares, and just overall unstable machine. I dont have this issue with my other machines on Trex or Phoenix…
i hope they update it soon, so it can be stable!
I cant set my machine for unknown reason to mine on MiniZ. looks like this miner is also giving good MH for 3060 cards, but, i tried everything! it wont start to mine at all…
So, i will wait for better days on lolminer…
by they way - using mem 1450 and core 2150
corelock is very efficient, and temp are under 50c…with 35-40% fans!
PL is like 105 at most…
driver: 460.39
tried 470…but i guess for my cards, that is the best one so far
WIll try 1150 again next time on lolminer. Haven’t been able to get dual mining on seperate pols with nbminer, but the results are good. Got 35.4-35.7 MH/s at around 104-110W. Did end up pushing the clocks a bit more (c 1580, m 2100, pl 110)
but from time to time , the hashrates are all over the place
gpu1 - goes missing lets say, after 30 sec appears again, after that gpu3 does the same thing
or gpu goes from 39 to 23 and up back to 29…
its not doing it constantly - but its frustrating big time!
it used to be very stable before - lower hashrates - yes, but stability was much much better!
Ok, which one?
I think i tried 470.(dont rememmber) but it wont start mining, even after 20 restarts!
Now, after power outage, wont start mining with 460.39…i dont know, very weird and annoying situation.
I went on the lolminer github place, and there was no much info on what is the best driver…
i tried your driver, wont start mining at all
tried with older hiveos and newer hiveos versions
i finally managed to make it work - used the 460.32.03 and lolminer 1.33
those are the results and we will see about stability
i hope it will run better and i guess i dont have another choice but to wait for better lolminer updates
try to increase mem to between 2400 and 2600… also try nvidia driver 495.44 and PL 120… see if it improves… you have non-lhr cards and should get around 40mh on each … on my lhr v2 3060 zotac, I’m getting 36mh on each