3060 Ti lhr very inconsistent mining with Nicehash.

tyruswoll

n00b
Joined
Sep 6, 2021
Messages
2
Recently bought a brand new RTX 3060 Ti lhr dual. Been mining on same rig before with a GTX 1070. Was getting very consistent results with nicehash quickminer. switched the 1070 out with the 3060, now i only get $2.80 avg with quickminer. where nicehash says i should be getting up to 4.87 a day.

Tried switching over to the nicehash non-quickminer version. getting up to $5 dollars on actual btc earned. but my problem is its not consistent. Miner switches off often. gets 8 dollars actual one minute and 1 dollar the next. Are there any fellow miners on nicehash with a 3060 ti lhr what algorithms are you using with which overclock settings.
 

MrGuvernment

Fully [H]
Joined
Aug 3, 2004
Messages
20,251
The rates vary, nothing is 100% consistent, You are selling mining power via nicehash, it is about demand and how many other people are selling at the same time. I can go from making $4 one second to it spiking upto $10 for a second and back down. If you want consitent 100% same rates, do not mine with nicehash, get an eth miner and join a pool and move your ETH to an exchange or something to trade.
 

noko

Supreme [H]ardness
Joined
Apr 14, 2010
Messages
6,542
Nicehash does contracts, people pay for hash power for a given coin, coins they want, coins they think have a future upswing, whatever. It may not be profitable for you but for NiceHash they will fulfill the contract, so you will see time spent on low profitability coins or Algorithms for those coins which you are providing your hardware to them to do. Now in Nicehash, have not used it in years, you can decide which algorithm(s) you will mine, it could be just Ethash or a combinations of Ethash and Autolykos2 or more algorithms. Just select the higher paying algorithms. I had times when my cards would have been idle since there was not sufficient contracts. Anyways the switching from one algorithm to another is time not mining which cuts into your profits.

I recommend going to pools where you decide what coin you want to mine, when and where they get sent, most don't have minimum balance that takes long to mine for. I just made a lot more profits going that route. I even solo mined Pirate and made out big when the going was easy. Many options. Plus tweaks become available for miners to try and to increase rewards. It turns out to be a rather interesting strategy game for resources, assets (hardware upgrades, purchases, configuration of spaces, power, cooling), which coin to mine as well as best pools. Not to mention long term short term goals, trades etc. My thoughts is why buy when you can mine pennies on the dollar at times?
 

noko

Supreme [H]ardness
Joined
Apr 14, 2010
Messages
6,542
whats the best eth miner? I havnt tried that so far.
For Nvidia, I've found T-Rex slightly better than the others. For AMD, TeamRed has some algorithms it hits out of the park and is my favorite miner. You also have Nano miner which works for both AMD and Nvidia and easy to setup to Nano pool (great pool), miner made by them but will also work with other pools.
 

sgrinavi

[H]F Junkie
Joined
Jan 18, 2009
Messages
10,230
whats the best eth miner? I havnt tried that so far.
3060ti LHR is actually worse than a 1080ti on ETH, the best you can do is use NBminer, I've hit as high as the low 40's using it. NBminer has bypassed the LHR lock somewhat.
 

arnemetis

2[H]4U
Joined
Aug 2, 2004
Messages
3,545
If you're using nicehash, grab this instead https://github.com/angelbbs/NiceHashMinerLegacy install msi afterburner if not installed already, and fire it up. You can specify additional algos such as ergo, which will be much more profitable on an lhr card than eth. My 3060 ti spends most of it's time on ergo, my 3070 ti on raven. They will sometimes switch to other algos depending on nh's current profitability. You can set different overclocks per algo and even per specific miner, which is great as for example ergo uses less power. Once your settings are dialed in run the benchmarks, then the program will take care of everything for you.
 

xenium

[H]ard|Gawd
Joined
Aug 15, 2006
Messages
1,734
3060ti LHR is actually worse than a 1080ti on ETH, the best you can do is use NBminer, I've hit as high as the low 40's using it. NBminer has bypassed the LHR lock somewhat.
Not really, and certainly not for much longer. 1080ti's are losing hashrate over time on eth (all pascal cards are). Theyre about even now, but the 1080ti's are almost twice the power draw. On any other coin the LHR 3060ti is a much better option.

3060ti - 40-42mh/s @120w
1080ti - ~43mh/s @ 200w+
 

arnemetis

2[H]4U
Joined
Aug 2, 2004
Messages
3,545
Not really, and certainly not for much longer. 1080ti's are losing hashrate over time on eth (all pascal cards are). Theyre about even now, but the 1080ti's are almost twice the power draw. On any other coin the LHR 3060ti is a much better option.

3060ti - 40-42mh/s @120w
1080ti - ~43mh/s @ 200w+
I think you're using old 3060 ti numbers, a 3060 ti lhr card will do 30mh @130w.
 

xenium

[H]ard|Gawd
Joined
Aug 15, 2006
Messages
1,734
Ah yes I forgot about that, sorry.
No worries. I actually have a few lhr 3060ti's on Erg right now, but theyre a bitch and a half to tune, its just really tricky getting them not to hit the LHR trigger point. But it was worth it as theyre all around ~100w or so, vs the 120-130w on Eth.

Fwiw Ive also heard some rumblings that other, further advances are being made with the LHR limitation. I think it's just a matter of time personally, time will tell if/when something is ever released publicly though. I wouldn't doubt it if some farms already have a privately funded 80-100% miner running right now. Theres so much money in the mining industry it's absurd, and it's crazy how lucrative it is to crack LHR right now.
 

sgrinavi

[H]F Junkie
Joined
Jan 18, 2009
Messages
10,230
Not really, and certainly not for much longer. 1080ti's are losing hashrate over time on eth (all pascal cards are). Theyre about even now, but the 1080ti's are almost twice the power draw. On any other coin the LHR 3060ti is a much better option.

3060ti - 40-42mh/s @120w
1080ti - ~43mh/s @ 200w+

Yeah? Are you running two or more 3060tis in a single rig? I am, I can get 40ish when I run a single LHR, but when you add the 2nd card the hash drops to the point where they're not efficient enough to make them worth the price tag on ETH. The 1080ti on that rig is running at 65% PL

1631103766193.png
 

xenium

[H]ard|Gawd
Joined
Aug 15, 2006
Messages
1,734
Yeah? Are you running two or more 3060tis in a single rig? I am, I can get 40ish when I run a single LHR, but when you add the 2nd card the hash drops to the point where they're not efficient enough to make them worth the price tag on ETH. The 1080ti on that rig is running at 65% PL


Yes I am, and that is weird, that looks like texhbook LHR behavior (normal ~60mhs cut in half). I tested eth on mine at 1300mhz locked core, +2000 mem (1000 windows) and saw them all bounce around 40-42mh/s.

Also you can pick up some hashrate on that 1080ti, memory utilization is only 79%. It sounds counterintuitive but dropping the memory clock and using better timings (ethlargement pill) will actually gain hash, until its near 100% MUtl.

This was from helping someone tune their 80ti a few months ago, so its probably a little lower hash now, but still shows youre leaving a bit of hashrate on the table for sure.
1080ti.png


+125 core, +400 mem (200 windows), 180PL. Note the mem utilization at 99%..
 

sgrinavi

[H]F Junkie
Joined
Jan 18, 2009
Messages
10,230
Yes I am, and that is weird, that looks like texhbook LHR behavior (normal ~60mhs cut in half). I tested eth on mine at 1300mhz locked core, +2000 mem (1000 windows) and saw them all bounce around 40-42mh/s.

love to see screen dumps when you get a chance.

Also you can pick up some hashrate on that 1080ti, memory utilization is only 79%. It sounds counterintuitive but dropping the memory clock and using better timings (ethlargement pill) will actually gain hash, until its near 100% MUtl.

This was from helping someone tune their 80ti a few months ago, so its probably a little lower hash now, but still shows youre leaving a bit of hashrate on the table for sure.View attachment 393060

+125 core, +400 mem (200 windows), 180PL. Note the mem utilization at 99%..

Interestingly those settings for the 1080ti didn't have the desired impact. Basically just lowered the hash a tiny amount. It's a gigabyte too, but I think I lost the silicon lottery on that one as it's never really been a performer. I'll probably yank it when I get another 30xx. My other 1080ti runs a bit better, but it has it's own set of issues.
 

xenium

[H]ard|Gawd
Joined
Aug 15, 2006
Messages
1,734
love to see screen dumps when you get a chance.



Interestingly those settings for the 1080ti didn't have the desired impact. Basically just lowered the hash a tiny amount. It's a gigabyte too, but I think I lost the silicon lottery on that one as it's never really been a performer. I'll probably yank it when I get another 30xx. My other 1080ti runs a bit better, but it has it's own set of issues.
The key with the 1080ti is running tighter timings and lower mem clock (ethlargement pill or other methods), but yeah could just be shitty luck too

Youre in luck I saved a shot from some testing actually. You also might be having issues with such a mixed rig (pascal and lhr ampere), but thats just a thought *shrug*

eth lhr.png


That was using a PL to keep the core clock down, but iirc locking it to 1300-1400 had the same effect/results.
 

sgrinavi

[H]F Junkie
Joined
Jan 18, 2009
Messages
10,230
The key with the 1080ti is running tighter timings and lower mem clock (ethlargement pill or other methods), but yeah could just be shitty luck too

Youre in luck I saved a shot from some testing actually. You also might be having issues with such a mixed rig (pascal and lhr ampere), but thats just a thought *shrug*

View attachment 393107

That was using a PL to keep the core clock down, but iirc locking it to 1300-1400 had the same effect/results.

I run the pill, it helps, but not much change with your settings.

Nice on the 3060ti rig, thanks for posting. I had them running with two other 30xx gpus and they did the same thing. Maybe I'll turn the 10xx gpus off and see what happens
 
Last edited:
Top