More RTX 3060 Nicehash overclocking

RTX 3060 Nicehash overclocking - 25% improvement

This is part 2 from my previous post on RTX 3060 Nicehash overclocking settings. I don’t want to edit the previous article because the content still stands to be accurate for the hash rate I achieved. However, I have since learned even more about the card and managed to improve my Nicehash quick miner hash rate by a further 10%!

If you want to see the first/part1 post, you can do so here: RTX 3060 Nicehash mining overclock settings
And if you are looking for my post (and the download) on unlocking the hash rate with the 470.05 driver, that’s here: Unlock RTX 3060 mining hash rate.

My previous settings to acheive 44 MH/s used the following sentiment:

  • Lower the core clock rate
  • Lower the power limit(%)
  • Increase the memory clock rate

But I have since made some overclocking improvements and some additional changes to improve the longevity of the card. So much so that my overall uplift from stock (using the 470.05 driver) is sitting at around 25%! I am now able to reach 48MH/s with the DaggerHashimoto algorithm.

RTX 3060 Nicehash overclocking - 25% improvement
Behold, 48 MH/s on an RTX 3060 (the card which NVidia does not want you to mine on)

RTX 3060 overclocking with MSI Afterburner

Your millage may vary with your individual card, but my improved settings to reach higher quickminer hash rates are as follows:

  • Power limit: 75%
  • Core Clock: -500 MHz – as low as it will go
  • Memory Clock + 1300 MHz
  • Fan Speed: Auto – with a more aggressive fan curve, see below…
RTX 3060 Nicehash overclocking - improved hash rate and fan curve
The MSI Afterburner settings to achieve 48MH/s on a RTX 3060

Keeping the GPU safe

RTX 3060 Nicehash overclocking - more aggressive fan curve for low VRAM temperatures.
A more aggressive fan curve to cool VRAM

I had been warned by a kind comment on my last post that whilst my GPU’s core clock temp might be low, the overclocked VRAM temp might be extremely high and limiting the performance of the card. What’s more it would not be conducive to a long living graphics card.

I use my profiles in MSI Afterburner to switch back to stock settings when I want to play some video games – I am not a dedicated miner. So killing my card prematurely is not something I want to proceed in doing. I have chose to keep my fan on auto, to spare my ears, but applied a more aggressive fan curve to help keep the card temps down.

My particular Gigabyte card does not have a dedicated VRRAM temperature sensor – or at least it is not recognized by GPU-Z or HWinfo. Therefore, my only hope is to keep the average temperate of the card under control.

(If you are going to follow this step, remember to tick the box to ‘Enable user defined software automatic fan control’)

12 Replies to “More RTX 3060 Nicehash overclocking”

  1. What is the power limit as value? I want to apply these in HiveOS for Conflux mining.

    I’m getting a lot of rejected values, and there are basically no mining OC guides for the 3060 :/.

    Thanks in advance

    1. The RTX 3060 has a TDP of 170W. I am running mine at 70% power limit (according to MSI afterburner), so it is limited to ~120W. You can only set absolute values in HiveOS OC profiles, not percentages. SO you would need to set: 120.00. Hope this helps!

      1. That helps out a lot! thank you.

        noob question, but does lowering Core clock reduce temperature?

        my gpus on hive OS are sitting around 55~65C, with fans at 40~50% speed, and my core clock at +100, and memory around 2400. if I reduce it to -200 or something will that help with temperature?

        1. Lowering the core clock can help to reduce the card’s overall temperature. But by increasing your mem-clock, you will be increasing the temperature of your VRAM – my Gigabyte RTX 3060 does not have a vram specific temp sensor, so I increased my fan speed with a more aggressive fan curve to err on the side of caution and extend the longevity of my GPU. Of course, there are a lot of environmental factors to include in that decision. I run my GPU in an ATX case with a typical room ambient temp of ~21C.
          The best way to see what impact each overclock option does is to open up GPU-Z sensors view and MSI afterburner (in a Windows environment), this will allow you to learn the direct impact overclock changes have on: board wattage, temp, and performance limiters.

    2. My Palit 3060 Dual (non-OC) achieved 49,59 MH/s on Hive OS with this setting fan 80, core 300, mem 2600, pl 110.

  2. hey tom, talk to me about the settings, i have a 3060 ghost oc, which currently runs 47 ~ 48mhz, my configuration is a 105w power limit (it stays at maximum 108w) core -420 mclock +1000 and fans at 60% and the temperatures are at maximum 55c, I believe it was the best optimized configuration I got, what do you think? have you changed any settings?

    1. I am more or less using the same settings and seeing the same results. I think we have reached peak hashrate for the RTX3060 for these particular drivers. The only further testing a am planning to test the card in HiveOS (because I’m seeing a lot of comments related to this) and in NicehashOS.

    2. Thats a good MH rate I have Asus 3060 Dual Fan non OC and I am averaging 46-47.8, I am happy enough with it and system is left on 24/7. Temps are 56/57 w fan at 74 and power limit at 69 and mem slider at +1077. Power draw seems to be steady at 111 power efficiency or slightly less. Gonna try Axel settings above , that’s pretty good as well

  3. No matter what I do I cant get my zotac twin edge oc 3060 to get more than 26 hashrate. My 1660 supers get about 30-31 so i really don’t understand what I’m doing wrong. I’m using phoenixminer on HiveOS

  4. To get full hashrate 3060 needs 370.04 Nvidia driver, needs to be plugged in main pci-e and have a screen connected. If not it will give half.

    I get 49,5 hashrate on my 3060 vision with the settings in this article. I tuned it down to 48,5 and got down the temp to 51 degres C.

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.

%d bloggers like this: