miniZ v1.8y4rc2 – beta version- is out!

Hi everyone,

Thank you for all your support and feedback. The new release miniZ v1.8y4rc2 (beta) is out.

We’ve been working with the RTX 3060, RTX 3070 Ti, and the RTX 3080 Ti. We manage to improve performance for LHR GPUs, on ethash mining.

Although it is not yet perfect, we hope this version will offer improvements for all of you.

Also, with this version we improved --tune option. Now the option is more flexible and you will be able to better tune your GPU/Rig. (Read below/check FAQ)

Below, you can find useful information about our testing settings. For more information check our Usage/ FAQ pages.

*** Download miniZ version v1.8y4rc2 here. ***

Changelog:

  • Added support for RTX 3070 Ti and 3080 Ti LHR GPUs, on ethash.
  • Improved performance for mining locked LHR GPUs while mining ETH.
  • Improved --tune option for better GPU tuning.
  • Small improvements for RTX 3060 on 144,5.

Let us know how it goes!

Next, we share some information about drivers, OCs, and --tune function. *

Driver
Because it is not very efficient or practical to work on all drivers, we focused on just a few but it’s likely that with other driver versions it will also work. The following ones were the most tested:

  • Linux: 460.39, 470.63, 470.74;
  • Windows: 461.40, 472.12. (It doesn’t work with 471.96).

The older driver 460.39 (Linux)/461.40 (Windows) doesn’t work for all 3060s. You can always try it out, but if it doesn’t work (or if you have other RTX 30XX, RTX 3080 Ti for example) then you have to install a more recent driver.

OCs
We also leave here information regarding the OCs we used for testing (driver 470.74). Note that you may need to adjust a bit for your system.

  • 3060: core: -150, mem:+2700 (Linux)*, PL=100-110 W; ~36.5 MH/s
  • 3070 Ti: core: -200, mem: +3700 (Linux)*, PL=185 W; ~55 MH/s
  • 3080 Ti: core: +0, mem: +2900 (Linux)*, PL=265 W. ~82 MH/s

* On Windows, memory clock with MSI Afterburner should be half of the indicated value. For example, for the 3080 Ti you need to set +1450 for memory clock in MSI Afterburner. Note that you may need to adjust a bit for your system.

--tune
We modified this option to allow more flexibility when tuning the GPU/Rig. It will work with different parameters (not compatible with the previous version) that you will be able to adjust for your GPU. For more information see FAQ.

The default values are not yet optimal for everyone. We suggest you to play around with i and d starting with small variations around the defaults.

--tune=iN1dN2 where i stands for intensity and d stands for drag. N1and N2 are numerical values.

  • Intensity works quite similar to the previous value that you were using before. N1 can take values in the interval [0, 64].
  • Drag will balance intensity. N2 values can vary in the interval [0, 100].

Note: If you increase PL, usually you’ll need to decrease i value.

More information about --tune usage can be found in FAQ page.

 

*** Thank you all for the feedback! ***

 

For additional information check our Usage or FAQ pages.

Engage, feedback and support

Because your feedback is very important, miniZ kindly encourages you to use the comment box below for general questions and feedback.

Certainly we will strive to help you with any doubt and provide a solution to any problem you may encounter. In case you need to reach us, please check the Support page.

Follow us
On Twitter
Our thread on bitcointalk.org

We wish you a fast and friendly mining experience!

Comments

  1. 3070ti says:

    good! I’m waiting update on hiveos.

    • OWL_DV says:

      You can manual update from SSH

      • Manuel says:

        Hi, can you tell me the easiest way for this before i have to spend much more time in searching… ? Thanks!

        • miniZ says:

          Hi Manuel,
          You can check here, for some suggestions on how to update in Hive OS.

          You can download from our website (or from our github if it will be easier for you).
          The following should work:
          cd /tmp; wget -q https://github.com/miniZ-miner/miniZ/releases/download/v1.8y4rc2/miniZ_v1.8y4rc2_linux-x64.tar.gz; tar -xf /tmp/miniZ_v1.8y4rc2_linux-x64.tar.gz miner stop; mv /hive/miners/miniz/1.8y4/miniZ /hive/miners/miniz/1.8y4/miniZ.bak; cp /tmp/miniZ /hive/miners/miniz/1.8y4/miniZ; miner start

          Cheers

  2. zz says:

    v1.8y4rc2 WIN10 Could not started

    • miniZ says:

      Hi zz,
      Thank you for your message!
      We have just replaced the file. Could you please download the new file and try again?
      Cheers

  3. 3080ti says:

    v1.8y4rc2 WIN10 Could not started,”time is out”

  4. 3070ti says:

    my 3070ti can’t use mem 3700. when I put 3700. low hash rate.

    RTX 3070 Ti  100% [41°C/60%]  35.11(48.60)MH/s 105(166.0)W clk=1725MHz mclk=11102MHz MH/W=0.29
    3070ti -200 / +3200 185w --tune=i15d60

    RTX 3070 Ti  100% [50°C/60%]  56.01(50.70)MH/s 185(172.5)W clk=1290MHz mclk=10852MHz MH/W=0.29

    and

    3060 lhr v2 29~35HM/s -150 +2700 110w –tune=i15d55  -> can’t reach 36HM/s…
    RTX 3060     ` 100% [57°C/85%]  35.12(34.68)MH/s 110(105.2)W clk=1417MHz mclk=8650MHz MH/W=0.33

    • miniZ says:

      Hi 3070Ti,
      The values we share worked reasonably well for us but we keep working to make this more stable as we are not fully satisfied with it, yet.
      If you can’t use mem=+3700 then you need to define a mem clock that best suits your system. The goal is to set a value that runs stable and delivers the highest hashrate.
      However, 56MH/s is not bad at all if it runs stable 🙂
      You can also try to lower the power, specially if it is not running stable for you.
      Thanks for using miniZ.
      Cheers

  5. zz says:

    Is the format right –tune i64d75

    • miniZ says:

      Hi zz,
      The format --tune i64d75 is correct.
      You can read some examples in How to use --tune for RTX 30XX (ethash)? @ FAQ page.

      i64 is a bit extreme value, we think that the miner will not be able to get rid of the lock. From our tests the working values are usually lower than 20, for the latest driver.

      However, let us know how it goes for you.
      Cheers

  6. asr says:

    DAG generation makes all subsequent shares stalev1.8y4rc2 WIN10, EVGA 3060 LHR V2, mining on NiceHash[ 0d 0h19m10s] 0>RTX 3060 100% [58 C/72%]* 34.83(34.76)MH/s 101(100.8)W clk=1438MHz mclk=8642MHz MH/W=0.35[ 0d 0h19m20s] 0>RTX 3060 100% [59 C/72%]* 34.85(34.77)MH/s 101(100.8)W clk=1438MHz mclk=8642MHz MH/W=0.35[ 0d 0h19m30s] 0>RTX 3060 100% [59 C/72%] 34.88(34.77)MH/s 102(100.8)W clk=1439MHz mclk=8642MHz MH/W=0.35[ 0d 0h19m40s] 0>RTX 3060 100% [58 C/72%]* 34.89(34.78)MH/s 100(100.7)W clk=1438MHz mclk=8642MHz MH/W=0.35[ 0d 0h19m50s] 0>RTX 3060 100% [58 C/72%]* 34.90(34.78)MH/s 101(100.8)W clk=1438MHz mclk=8642MHz MH/W=0.35[INFO ] GPU[0]: Generating DAG #0 [1023+15=1039MB][INFO ] GPU[0]: Generated DAG in 2.135[ 0d 0h20m00s] 0>RTX 3060 100% [56 C/72%]* 34.91(34.78)MH/s 91(100.5)W clk=1438MHz mclk=8642MHz MH/W=0.35[INFO ] Target set to 0000000061B5CB9E (512.42M)[ 0d 0h20m10s] 0>RTX 3060 100% [57 C/72%] 34.91(34.78)MH/s 92(100.4)W clk=1439MHz mclk=8642MHz MH/W=0.35[ 0d 0h20m20s] 0>RTX 3060 100% [56 C/72%] 32.24(34.37)MH/s 88(100.1)W clk=1439MHz mclk=8642MHz MH/W=0.34[ 0d 0h20m30s] 0>RTX 3060 100% [56 C/72%] 32.24(34.37)MH/s 86( 99.9)W clk=1439MHz mclk=8642MHz MH/W=0.34[ 0d 0h20m40s] 0>RTX 3060 100% [56 C/72%] 29.50(33.93)MH/s 85( 99.6)W clk=1439MHz mclk=8642MHz MH/W=0.34[ 0d 0h20m50s] 0>RTX 3060 100% [55 C/72%] 29.50(33.93)MH/s 85( 99.4)W clk=1439MHz mclk=8642MHz MH/W=0.34[ 0d 0h21m00s] 0>RTX 3060 100% [55 C/72%] 26.99(33.49)MH/s 84( 99.1)W clk=1439MHz mclk=8642MHz MH/W=0.34[ 0d 0h21m10s] 0>RTX 3060 100% [55 C/72%] 26.99(33.49)MH/s 84( 98.9)W clk=1439MHz mclk=8642MHz MH/W=0.34[ 0d 0h21m20s] 0>RTX 3060 100% [55 C/72%] 24.69(33.07)MH/s 84( 98.6)W clk=1439MHz mclk=8642MHz MH/W=0.34[ 0d 0h21m30s] 0>RTX 3060 100% [55 C/72%] 24.69(33.07)MH/s 84( 98.4)W clk=1439MHz mclk=8642MHz MH/W=0.34[ 0d 0h21m40s] 0>RTX 3060 100% [54 C/72%] 22.59(32.67)MH/s 84( 98.2)W clk=1439MHz mclk=8642MHz MH/W=0.33[ 0d 0h21m50s] 0>RTX 3060 100% [54 C/72%] 22.59(32.67)MH/s 84( 98.1)W clk=1439MHz mclk=8642MHz MH/W=0.33[ 0d 0h22m00s] 0>RTX 3060 100% [54 C/72%] 20.67(32.27)MH/s 84( 97.9)W clk=1439MHz mclk=8642MHz MH/W=0.33[INFO ] Target set to 00000000C36B973C (256.21M)[ 0d 0h22m10s] 0>RTX 3060 100% [54 C/72%] 20.67(32.27)MH/s 84( 97.6)W clk=1439MHz mclk=8642MHz MH/W=0.33[INFO ] Target set to 00000000C36B973C (256.21M)[ 0d 0h22m20s] 0>RTX 3060 100% [54 C/72%] 18.91(31.88)MH/s 84( 97.4)W clk=1439MHz mclk=8642MHz MH/W=0.33[WARNING] Stale share: Job not found.[WARNING] Stale share: Job not found.[WARNING] Stale share: Job not found.[WARNING] Stale share: Job not found.[WARNING] Stale share: Job not found.[WARNING] Stale share: Job not found.[WARNING] Stale share: Job not found.[WARNING] Stale share: Job not found.Same issue happens regardless of OC settings and tune; no auto recovery (no share gets accepted afterwards, only restarting miniZ solves the issue)

  7. OWL_DV says:

    HelloHow to direct download form WGET miniZ_v1.8y4rc2_linux-x64.tar.gz ?wget not work – https://miniz.ch/?smd_process_download=1&download_id=4580

  8. asr says:

    DAG generation makes all subsequent shares stale. v1.8y4rc2 WIN10, EVGA 3060 LHR V2, mining on NiceHash           

    [INFO ] GPU[0]: Generating DAG #0 [1023+15=1039MB]
    [INFO ] GPU[0]: Generated DAG in 2.135
    [ 0d 0h20m00s] 0>RTX 3060 100% [56 C/72%]* 34.91(34.78)MH/s 91(100.5)W clk=1438MHz mclk=8642MHz MH/W=0.35
    [INFO ] Target set to 0000000061B5CB9E (512.42M)
    [ 0d 0h20m10s] 0>RTX 3060
    100% [57 C/72%] 34.91(34.78)MH/s 92(100.4)W clk=1439MHz mclk=8642MHz MH/W=0.35
    [ 0d 0h20m20s] 0>RTX 3060 100% [56 C/72%] 32.24(34.37)MH/s 88(100.1)W clk=1439MHz mclk=8642MHz MH/W=0.34
    [ 0d 0h20m30s] 0>RTX 3060
    100% [56 C/72%] 32.24(34.37)MH/s 86( 99.9)W clk=1439MHz mclk=8642MHz MH/W=0.34
    [ 0d 0h20m40s] 0>RTX 3060 100% [56 C/72%] 29.50(33.93)MH/s 85( 99.6)W clk=1439MHz mclk=8642MHz MH/W=0.34
    [ 0d 0h20m50s] 0>RTX 3060
    100% [55 C/72%] 29.50(33.93)MH/s 85( 99.4)W clk=1439MHz mclk=8642MHz MH/W=0.34
    [ 0d 0h21m00s] 0>RTX 3060 100% [55 C/72%] 26.99(33.49)MH/s 84( 99.1)W clk=1439MHz mclk=8642MHz MH/W=0.34
    [ 0d 0h21m10s] 0>RTX 3060
    100% [55 C/72%] 26.99(33.49)MH/s 84( 98.9)W clk=1439MHz mclk=8642MHz MH/W=0.34
    [ 0d 0h21m20s] 0>RTX 3060 100% [55 C/72%] 24.69(33.07)MH/s 84( 98.6)W clk=1439MHz mclk=8642MHz MH/W=0.34
    [ 0d 0h21m30s] 0>RTX 3060
    100% [55 C/72%] 24.69(33.07)MH/s 84( 98.4)W clk=1439MHz mclk=8642MHz MH/W=0.34
    [ 0d 0h21m40s] 0>RTX 3060 100% [54 C/72%] 22.59(32.67)MH/s 84( 98.2)W clk=1439MHz mclk=8642MHz MH/W=0.33
    [ 0d 0h21m50s] 0>RTX 3060
    100% [54 C/72%] 22.59(32.67)MH/s 84( 98.1)W clk=1439MHz mclk=8642MHz MH/W=0.33
    [ 0d 0h22m00s] 0>RTX 3060 100% [54 C/72%] 20.67(32.27)MH/s 84( 97.9)W clk=1439MHz mclk=8642MHz MH/W=0.33
    [INFO ] Target set to 00000000C36B973C (256.21M)
    [ 0d 0h22m10s] 0>RTX 3060
    100% [54 C/72%] 20.67(32.27)MH/s 84( 97.6)W clk=1439MHz mclk=8642MHz MH/W=0.33
    [INFO ] Target set to 00000000C36B973C (256.21M)
    [ 0d 0h22m20s] 0>RTX 3060 ` 100% [54 C/72%] 18.91(31.88)MH/s 84( 97.4)W clk=1439MHz mclk=8642MHz MH/W=0.33
    [WARNING] Stale share: Job not found.
    [WARNING] Stale share: Job not found.
    [WARNING] Stale share: Job not found.
    [WARNING] Stale share: Job not found.
    [WARNING] Stale share: Job not found.
    [WARNING] Stale share: Job not found.
    [WARNING] Stale share: Job not found.
    [WARNING] Stale share: Job not found.

     

    Same issue happens regardless of OC settings and tune; no auto recovery (no share gets accepted afterwards, only restarting miniZ solves the issue)

    • miniZ says:

      Hi asr,
      Thank you for the feedback.
      We had a look at this. It seems that NiceHash ocasionaly sends a null header to the miner.
      So far, we were unable to deal with this. Do you only observe this with miniZ?
      Cheers

  9. OWL_DV says:

    Colleagues!
    Please give me the optimal overclocking and miner settings for 3060 lhr v2.
    I have 15 pieces, Samsung memory.

    • Alvaro Lordelo says:

      Hi friend, so check first if it’s device 2504, if so..I’ll recommend you to try – CC 1450 MC 2350Fan 80Power Limit 118—-my second option is CC 1250 MC 2600 Fan 80 PL 118

    • miniZ says:

      Hi OWL_DV,
      As we mention here, the following OCs worked well with our 3060:
      core: -150, mem:+2700 (Linux)*, PL~100W; ~36.5 MH/s
      Cheers

  10. JohNY says:

    how to upload last beta version to hive OS?

  11. mthoring says:

    rc2 looks really promising on 3060 lhr v2. Currently at ~37mh/s@115w (-150c, +2850mem).Quickly locked @120w, really hoping it will be stable at 115w or 110w.The highest I was able to reach with nbminer was 34mh/s using either lhr-modes. And it kept locking at random intervals.

    • Raul says:

      37 mhs on 3060 LHR regular?I need to see it. Please, post a screenshot. Wiith your OC I get locked every 20 seconds: [TRACE ] GPU[2]: Searching lock… speed=22.1/48.3[1,4864.25,2.070,0.85] My actual OC, 35.2 Mhs @105w stable and realiable:CC 1450Mem 2575PL 120w

    • 3060 LHR v2 says:

      Could you share your NVidia driver version?Miniz parameters?

    • miniZ says:

      Hi mthoring,
      Thank you for the feedback.
      We keep working on this, trying to make it more stable.
      Cheers

  12. iMine says:

    Works well with my MSI 3060 Ti (Hynix) with locked core clock and mild OC. Reached a peak >41 MH/s, the best I got across all miners I tried. However the hashrate collapses after 50 min (almost exactly, could repeat it twice). What’s happening?

    • miniZ says:

      Hi iMine,
      Thank you for your feedback.
      We’re working to make the miner more stable. We’re still trying to understand this better.
      Cheers

  13. RSG says:

    i cant get the lhr tune to work for my v2 3060./..

  14. 3070 Ti says:

    EVGA 3070 Ti works stable on windows @ 54~55MH/sThx miniZ

  15. Vit says:

    Manually updated to HiveOS. Catastrophic, Performance much worse than rc1, always locked, up to 30MH/s and crashing down. Is anyone sucesfull with this update?Could you share HW and OC settings?Rc1 works for me core -200 mem 2650 tdp 115wSame settings for rc2 is crappy in my case

    • Vit says:

      4x Gainward 3060 Ghost (LHR V2) driver 460.91, HiveOS 1.8yrc1 – approx 33.5-35.5MH/s / per card OC -200 2650 115W 1.8yrc2 unusable for me, can you suggest me better settings?

      • miniZ says:

        Hi Vit,

        The following OCs work well for us:
        3060: core: -150, mem:+2700 (Linux)/+1350 (Windows), PL=~100 W; ~36.5 MH/s

        If you lower power a bit more, does it help?
        Cheers

    • miniZ says:

      Hi Vit,
      Thank you for letting us know. We are sorry to hear that it is not working well for you.
      Have you tried to lower the power a bit more? And have you tried to use tune parameters?
      Check FAQ page, How to use --tune for RTX 30XX (ethash)?.
      Cheers

  16. Zotac OC Twin Edge 3060 12gb says:

    3060 LHR v2 (Zotac OC TWin Edge 12) report:OC CPU -150, MEM +1200, PL 70%Tune i15d55Starts at ~27.5 MH/s and falls to around 25 MH/s, then drifts around from 30-35TREX stable at 34I’ll keep MiniZ up for a while and see if the average increases

    • miniZ says:

      Hi Zotac OC Twin Edge 3060 12gb,
      Thank you for the feedback.
      It looks that the lock is appearing too much, try to lower PL a bit more.
      Let us know how it goes.
      Cheers

  17. Spoinky says:

    Hi, Thank for you great work. With miniZ (v1.8y4rc2), i manage on HyveOs and RTX 3060 : 3>RTX 3060 100% [50°C/ 0%] 40.74(40.64)MH/s 110(109.8)W clk=1777MHz mclk=7501MHz MH/W=0.37cpu focused options : –mode3 80 –tune3 37But after like 15min, miniZ restart with in between this error :GPU[3]: CUDA error 77 ‘an illegal memory access was encountered’ in line 878Best Regards.Ps : Merci poust ce super soft !

  18. Alex says:

    Hi support, how to disable logging in file? I cant find parameter for this. (linux) Regards

    • miniZ says:

      Hi Alex,
      miniZ does not log to file by default.
      To disable logging in file do not specify --log or --logfile in the command line.
      Let us know if this is what you mean by “disable logging in file”.
      Cheers

  19. Hsnk says:

    Hi

    I have 3060 and I want to crash driver for fan glitch, so how can i config ? Is that possible with —tune ?

    • miniZ says:

      Hi Hsnk,
      There is no recipe to cause this with miniZ.
      We understand why you want to do it, however, we do not advise this.
      Usually the driver can crash if the OC settings are too extreme.
      We are working to try to make the miner more stable.
      Cheers

  20. 3060 LHR V1 says:

    v1.8y4rc1 was much stable with my cards. Now my hashrate on my rig is 308 and was 317MHs How to downgrade to previous version?

    • miniZ says:

      Hi 3060 LHR V1,
      Sorry to hear that you find the previous version better. Have you tried to play around with --tune options?
      You can check the FAQ page for more information. Do pay attention that tune changed with the latest version, so the values to set will be different with v1.8y4rc2.
      To downgrade you only need to start the previous version. You may need to download it again if you have deleted the file already.

      Are you using Hive OS? To update you can see some suggestions from other comments here:
      https://miniz.ch/2021/10/06/miniz-v1-8y4rc2-beta-version-is-out/#comment-10520

      You can download from our website (or from github if it will be easier for you).
      The following should work:
      cd /tmp; wget -q https://github.com/miniZ-miner/miniZ/releases/tag/v1.8y4rc2; tar -xf /tmp/miniZ_v1.8y4rc2_linux-x64.tar.gz miner stop; mv /hive/miners/miniz/1.8y4/miniZ /hive/miners/miniz/1.8y4/miniZ.bak; cp /tmp/miniZ /hive/miners/miniz/1.8y4/miniZ; miner start

      Cheers

  21. dazexy says:

    Greetings!

     

    Pls. Anybody can tell me how to insall back (downgrade) to v1.8y4rc1?  I cant choose from list after hiveos upade.

    • miniZ says:

      Hi dazexy,
      Sorry to know that you need to downgrade. Have you tried to play around with --tune options?
      You can check the FAQ page for more information. Do pay attention that tune changed with the latest version, so the values to set with tune will be different in v1.8y4rc2.
      Check here, for some suggestions on how to update/downgrade in Hive OS.

      You can download from our website (or from github if it will be easier for you).
      The following should work:
      cd /tmp; wget -q https://github.com/miniZ-miner/miniZ/releases/tag/v1.8y4rc2; tar -xf /tmp/miniZ_v1.8y4rc2_linux-x64.tar.gz miner stop; mv /hive/miners/miniz/1.8y4/miniZ /hive/miners/miniz/1.8y4/miniZ.bak; cp /tmp/miniZ /hive/miners/miniz/1.8y4/miniZ; miner start

      Cheers

  22. Stevem2500 says:

    Rc2 lowered my hash rates on my 3060 v1 and significantly lowered my hash rates on my 3070 v1’s.  Rc1 was solid.  Please revert back for hiveOS!

    • miniZ says:

      Hi Stevem2500,
      Thank you for the feedback.
      Sorry to know that you need to downgrade. Have you tried to play around with --tune options?
      You can check the FAQ page for more information. Do pay attention that tune changed with the latest version, so the values to set with tune will be different in v1.8y4rc2.
      Check here, how to update/downgrade in Hive OS.

      You can download from our website (or from github if it will be easier for you).
      The following should work:
      cd /tmp; wget -q https://github.com/miniZ-miner/miniZ/releases/tag/v1.8y4rc2; tar -xf /tmp/miniZ_v1.8y4rc2_linux-x64.tar.gz miner stop; mv /hive/miners/miniz/1.8y4/miniZ /hive/miners/miniz/1.8y4/miniZ.bak; cp /tmp/miniZ /hive/miners/miniz/1.8y4/miniZ; miner start

      Cheers

  23. HiveUser says:

    Hi adminCan you please repost ideal OC and Driver version for RTX3060 (V1) with Linux?There are alot of posts and couldn’t find latest information.Does it have potential of more than 40MH/s?

  24. VICTOR HUGO PILASTRI says:

    3060: core: -150, mem:+2700 (Linux)*, PL=100-110 W; ~36.5 MH/s3070 Ti: core: -200, mem: +3700 (Linux)*, PL=185 W; ~55 MH/s3080 Ti: core: +0, mem: +2900 (Linux)*, PL=265 W. ~82 MH/s What tune did you use?

  25. PabloElRey says:

    Hello, I was mining with 5 RTX 3060 v1 at 40.5 MH until yesterday’s upgrade, which is now more unstable and has lower performance. What can I do?

    • miniZ says:

      Hi PabloElRey,
      Sorry to hear that the new version did not work well for you.
      Have you tried to adjust the tune values?
      You can check some information here.
      Also, maybe you could try to lower PL a bit more.

      However, if it does not work for you maybe it is best to use the previous version.
      We are working on a new version and trying to make this better.

      Are you using HiveOS? Check here some suggestions on how to update/downgrade in Hive OS:
      https://miniz.cc/2021/10/06/miniz-v1-8y4rc2-beta-version-is-out/#comment-10520

      You can download from our website (or from github if it will be easier for you).
      The following should work:
      cd /tmp; wget -q https://github.com/miniZ-miner/miniZ/releases/tag/v1.8y4rc2; tar -xf /tmp/miniZ_v1.8y4rc2_linux-x64.tar.gz miner stop; mv /hive/miners/miniz/1.8y4/miniZ /hive/miners/miniz/1.8y4/miniZ.bak; cp /tmp/miniZ /hive/miners/miniz/1.8y4/miniZ; miner start

      Cheers

  26. Aldo says:

    hi, i have a MSI 3060 v1, before the upgrade to 1.8y4rc2 with HiveOs i was stable at 40.2mhs (core 1500 mem 2300) on ethash. Now my hashrate goes up and down from 39.6 to 39.8. (i just notice more stable power consumption at 107-108w, while before with 1.8y4rc1 oscillate between 106 and 112w)Should

  27. Cri says:

    worst performance on hiveos with 3060v1

    • miniZ says:

      Hi Cri,
      Sorry to hear that it is not working for you.
      Try to lower PL to ~100W. If this doesn’t help, then maybe is best to use the previous version.
      Thanks for your feedback.
      Cheers

  28. Pilvx says:

    What tune you use?3060: core: -150, mem:+2700 (Linux)*, PL=100-110 W; ~36.5 MH/s3070 Ti: core: -200, mem: +3700 (Linux)*, PL=185 W; ~55 MH/s3080 Ti: core: +0, mem: +2900 (Linux)*, PL=265 W. ~82 MH/s

  29. Steven says:

    Hi, please add something like this in future release (GPU vendor device name detection).
    Great miner btw!

    https://gitlab.com/CalcProgrammer1/OpenRGB/-/blob/ce4af9726ee8ad5fe4c185f293de01eea1e33eb9/Controllers/GigabyteRGBFusionGPUController/GigabyteRGBFusionGPUControllerDetect.cpp

    Would make differentialization easier, instead of seeing “GTX 1080 TI” times 5 (having different vendors).
    A bunch of them exists under /OpenRGB/[Vendor]GPUController/[Vendor]GPUControllerDetect.cpp

    • miniZ says:

      Hi Steven,
      Thanks for your suggestion.
      The list isn’t complete but it’s a nice start.
      We’ll build this for the next release.
      Cheers

  30. alex1109 says:

    Hi together! I use a Palit 3080ti (not the OC one) with Nvidia driver 472.12 on Win10 pro. MiniZ_v1.8y4rc2 works pretty well between 75 to 89 MH/s. I set VRAM +1405 MHz and the rest I left at it was.Once a day I have to restart the miniZ. Unfortunally at a displaying hashrate about 2300 MH/s . Nice displaying but (of cause) it generates no profit – so it stucks an I have to restart the miner.

    • miniZ says:

      Hi alex1109,
      Thank you for your message.
      Do you mean that the miner starts reporting 2300MH/s and it hangs?
      It is a bit difficult to understand.
      Could you paste here your command line?
      Also, could you paste here the log output when it starts working funny?
      Cheers

  31. Steven says:

    (Bug) If i put computer to sleep while mining, and resuming in the morning, the “fee.server” will mine forever. BTCZ in this case, fee time 20.

  32. Konstantin says:
    Hi,nothing works on the recommended settings

    HiveOs , Nvidia driver 470.74


    Core -150 mem +2700 PL 100
    [TRACE ] GPU[05]: Searching lock …cy=11 speed=21.1/45.4[2,i15,d55.0,1.15,50,0.0] 4249
    [TRACE ] GPU[04]: Searching lock …cy=12 speed=22.9/46.5[2,i15,d55.0,1.15,50,0.0] 7779
    [TRACE ] GPU[01]: Searching lock …cy=13 speed=20.6/46.6[2,i15,d55.0,1.15,50,0.0] 4356
    [TRACE ] GPU[03]: Searching lock …cy=18 speed=23.0/46.4[2,i15,d55.0,1.15,50,0.0] 6715
    [TRACE ] GPU[06]: Searching lock …cy=7 speed=18.9/45.3[2,i15,d55.0,1.15,50,0.0] 7394
    [TRACE ] GPU[00]: Searching lock …cy=19 speed=21.8/45.6[2,i15,d55.0,1.15,50,0.0] 4167

    • miniZ says:

      Hi Konstantin,
      We are sorry to know that the settings are not working for you.
      It looks that the miner is not able to get rid of the lock.
      We’re working to improve stability.
      Thank you for the feedback.
      Cheers

  33. vijay says:

    Getting 39.5 MH/S EThash using the miniZ v1.8y4rc1 on my 6X 3060 Version 1 cards. Can this be further improved with miniZ v1.8y4rc2?pls advice.

    • StealYourBlock says:

      I’m getting similar results as you.3060v1, 460.39 driver, v1.8y4rc1, 39-40MH/s, -150 Core | 2300 Memory | 115w PL.Can we get more with v1.8y4rc2? I’m really curious please answer Miniz team.

      • miniZ says:

        Hi StealYourBlock,
        With your GPU, there will be no advantage if switching to v1.8y4rc2.
        But, 39-40MHs is very good!
        Cheers

    • Leki says:

      Same miner and same GPU but 5 of them. OC Core: -200 Memory:2300 PW:120 and 460.39 drivers on HiveOS. Getting 40.23 MH but 2% or more of rejected shares and is not stable. I have to restart at least once a day without OC and when DAG is loaded then I return OC.I tried with v1.8y4rc2 but I can’t get more than 36 MH with OC Core: -150 Memory:2700 PW:120 and 460.39 drivers on HiveOS. Tried tuning, mode… and was very unstable. If anyone can help I would be grateful.

      • miniZ says:

        Hi Leki,
        Thank you for the feedback.
        We suggest you to lower the power a bit more, to ~100W. I should be more stable.
        Let us know if this helps.
        Cheers

        • Leki says:

          Lowered PW to 105 and now is more stable. Now the efficiency is 99% even after 10 days, it used to be like that only in a few hours.

    • miniZ says:

      Hi vijay,
      Thank you for your message.
      It looks that it is working well for you.
      If you’re getting that hashrate with v1.8y4rc1 moving to v1.8y4rc2 will not bring advantage.
      Cheers

  34. Franco says:

    Please do the 40 MH/s for 3060 V1 cards with a newer driver (like 470.74 ) mine can´t work with the oldest driver

    • miniZ says:

      Hi Franco,
      Thank you for your message.
      Are you sure your 3060 is V1?
      To make 40MH/s you need the 460.39 driver.
      Cheers

  35. Oranged says:

    Hi, I’m having some trouble to understand how miniZ is working. I’m using a 3060 v2 in hiveOs with the following OC cc -150 m 2850 pl 120, n o additional argument set in the miner config. Running for 1 hour now, and unstable, going to 37mh then dropping to 31mh is it normal ? Is the miner auto-calibrating ? Thanks in advance !

    • miniZ says:

      Hi Oranged,
      It is normal with GPUs that have anti-mining mechanisms.
      But you may be able to improve stability and the hashrate if you decrease PL to ~100W.
      Let us know how it goes.
      Cheers

  36. styleshifter says:

    Are you still reading your thread and messages on bitcointalk?Some time ago I sent you a message there.If there are more convenient ways to contact you please tell me.

  37. Nikita says:

    Dears,

    Hive os

    GeForce RTX 3080 Ti 12053 MB ·
    MSI

    Micron GDDR6X · 94.02.71.40.11
    · PL 100 W, 350 W, 350 W

    as recomended

    core 150

    mem 2900

    pl 265

    in flight
    sheet

    ezil pool

    –tune
    i10d59

    And only
    60mh

    What I’ve
    done wrong?

    • miniZ says:

      Hi Nikita,

      You are right it should make ~80MH/s.
      Does the hashrate go up and down or is it running stable at 60Mh/s?

      You can try --tune i9d59 or --tune i10d58.
      Let us know if this helps.
      Cheers

  38. NucleaR-II says:

    Hello miniZ,are you working on a 100% LHR unlock for the 3080TI ?If so how long can this take ? THANKS

    • Nagrom says:

      Dude, If they could unlock it they would ! Stop asking this kind of question please.No time frame can be give as it is a work in progress and they have to constantly adapt.

      • miniZ says:

        Hi Nagrom,
        Thank you for understanding.
        Indeed, this has been quite challenging, making it hard to set deadlines.
        Cheers

    • miniZ says:

      Hi NucleaR-II,
      We’re working on stability improvements for all locked GPUs. 100% unlock is not possible at the moment.
      It is difficult to define a time range. We always hope it is very soon.
      However, sometimes, on longer tests we realize that is not good enough yet.
      Thank you for using miniZ.
      Cheers

  39. mxs says:

    ************ miniZ v1.8y4rc2 ************
    miniZ<ethash>[80:2:00.0:10225]: Selecting GPU#0[0] NVIDIA GeForce RTX 3060
    Driver: 472.12
    Algo: EQ[ethash] [smart-pers]
    Pool#0:  user[devilminers.test]
    server[ethereum.viabtc.com] port[3333] ssl[no] pers[ethash]
    Telemetry: [http://localhost:20000]
    Logging::  file[miniZ.log] period[10] delay[0]
    Temp. limit: [90 C]
    [INFO   ] Mining fee set to 0.75%
    [INFO   ] GPU[0]: Generating DAG #452 [4639+72=4712MB]
    [INFO   ] GPU[0]: Generated DAG in 12.928
    [ 0d 0h 0m24s] S:  0/0/0 0>RTX 3060  100% [43 C/30%]  56.55(56.55)kH/s 102(101.7)W clk=1957MHz mclk=7293MHz kH/W=0.56 [ethereum.viabtc.com]- 179ms
    [ 0d 0h 0m30s] S:  0/0/0 0>RTX 3060  100% [46 C/30%]  14.53(14.53)MH/s 108(108.3)W clk=1957MHz mclk=7293MHz MH/W=0.13 [ethereum.viabtc.com]- 179ms (100.00%) (0.00%)
    [ 0d 0h 0m40s] S:  0/0/0 0>RTX 3060  100% [48 C/30%]  14.19(14.19)MH/s 115(114.6)W clk=1945MHz mclk=7293MHz MH/W=0.12 [ethereum.viabtc.com]- 179ms (100.00%) (0.00%)
    [TRACE  ] GPU[00]: Searching lock …cy=4 speed=16.8/21.7[2,i19,d55.0,1.15,85,2.0] 112
    [ 0d 0h 0m50s] S:  0/0/0 0>RTX 3060  100% [47 C/30%]  14.19(14.19)MH/s 118(117.2)W clk=1945MHz mclk=7293MHz MH/W=0.12 [ethereum.viabtc.com]- 179ms (100.00%) (0.00%)
    [ 0d 0h 1m00s] S:  0/0/0 0>RTX 3060  100% [41 C/30%]  11.08(11.08)MH/s  79( 92.2)W clk=1777MHz mclk=7293MHz MH/W=0.12 [ethereum.viabtc.com]- 179ms (100.00%) (0.00%)
    [ 0d 0h 1m10s] S:  0/0/0 0>RTX 3060  100% [50 C/30%]  13.04(13.04)MH/s  91( 93.5)W clk=1944MHz mclk=7293MHz MH/W=0.14 [ethereum.viabtc.com]- 177ms (100.00%) (0.00%)
    [ 0d 0h 1m20s] S:  0/0/0 0>RTX 3060  100% [51 C/30%]  13.37(13.37)MH/s 108(100.9)W clk=1945MHz mclk=7293MHz MH/W=0.13 [ethereum.viabtc.com]- 177ms (100.00%) (0.00%)
    [TRACE  ] GPU[00]: Searching lock …cy=4 speed=21.5/41.2[2,i19,d55.0,1.15,85,2.0] 232
    [ 0d 0h 1m30s] S:  0/0/0 0>RTX 3060  100% [52 C/30%]  13.57(13.57)MH/s 116(106.2)W clk=1959MHz mclk=7293MHz MH/W=0.13 [ethereum.viabtc.com]- 177ms (100.00%) (0.00%)
    [ 0d 0h 1m40s] S:  0/0/0 0>RTX 3060  100% [44 C/30%]  13.57(13.57)MH/s  78( 93.2)W clk=1777MHz mclk=7293MHz MH/W=0.15 [ethereum.viabtc.com]- 177ms (100.00%) (0.00%)
    [ 0d 0h 1m50s] S:  0/0/0 0>RTX 3060  100% [50 C/30%]  11.60(11.60)MH/s  91( 93.6)W clk=1207MHz mclk=4633MHz MH/W=0.12 [ethereum.viabtc.com]- 177ms (100.00%) (0.00%)
    [ 0d 0h 2m00s] S:  0/0/0 0>RTX 3060  100% [55 C/30%]  13.66(13.66)MH/s 122(103.4)W clk=1962MHz mclk=7293MHz MH/W=0.13 [ethereum.viabtc.com]- 177ms (100.00%) (0.00%)
    [TRACE  ] GPU[00]: Searching lock …cy=4 speed=15.1/41.2[2,i19,d55.0,1.15,85,2.0] 51
    [ 0d 0h 2m10s] S:  0/0/0 0>RTX 3060  100% [54 C/30%]  15.14(15.14)MH/s 130(108.9)W clk=1963MHz mclk=7293MHz MH/W=0.14 [ethereum.viabtc.com]- 156ms (100.00%) (0.00%)

    • miniZ says:

      Hi mxs,
      Thank you for the feedback.
      Try to lower PL to ~100W. miniZ works best with lower PL.
      Let us know if this helps.
      Cheers

  40. Marcus Wittke says:

    Dear MiniZ Team,just to let you know, my Nvidia p106-100 micron memory (original Santa Clara, Ca. build) is working just fine ?? somewhere abouts 12.55-12.67 I/s 24.4x-25.67 Sol/s 95-110 Watts 0.24-0.25 Sol/Watt on Win 10I tried your overclocks, but does not work for me, could be the micron memory chipscheers und stay healthy,marxus

    • miniZ says:

      Hi Marcus Wittke,
      Thank you for the feedback.
      Did you try running the miner with --ocX?
      Let us know if this improves the hashrate by choosing a kernel mode different from the default.
      Cheers

  41. umenz says:

    hallo miniZ, i’m mining aion and the high stella share, how to fix ?, i use gpu k80

    • miniZ says:

      Hi umenz,
      What pool are you using, and which miniZ version?
      You can try adding --ocX to your command line. Maybe the miner is not selecting the optimal kernel for that GPU.
      Let us know how it goes.
      Cheers

  42. HWND says:

    3 Is the calculation force too low incorrect? 2060S, 32 forces, but 3090 is only 65 forces, how can that be, shouldn’t it be about 90 forces?

  43. HWND says:

    Please update it.Equihash 150,5 algo,2060S (32 Sol/S)3080TI (66 Sol/S) Too low!!

    • miniZ says:

      Hi HWND,
      What sort of values were you expecting?
      The 2070 does ~35 Sol/s, and the 2080Ti does ~60 Sol/s. So this values do not seem too bad.
      Cheers

  44. Miner says:

    Three issues

    1) miniZ doesn’t report to the log that URL was changed (RPC reconnect) It works good but this is important event and should be reported

    2) telemetry page (TCP port 20000) doesn’t have link to official miniZ web. Just add some MiniZ logo with a link.

    3) miniZ & RVN. I see many rejected shares. miniZ is not good for RVN. This example has “redirect”, port 3333 is redirected to custom TCP port (point 1). It can be used to see how shares are rejected too, during writing this post, it was like 7/1/3. I tested with GTX1070.

    ./miniZ --par=progpow --smart-pers --server=eu-de01.miningrigrentals.com:3333 --user=droidMiner.187431 --pass=TEST1
    • miniZ says:

      Hi Miner,

      1)Do you mean to log the pool information? You can add --show-server to your command line, to show server address. Or --extra that inclued this and more information. Read here for more options.

      2)We’ll add this for the next version.

      3)We’re not having this issue. With version 1.8y4rc2 this doesn’t seem to happen.
      Which version are you using?

      Thank you for your message.
      Cheers

  45. John Bacil says:

    I miss a command line parameter to debug communication with the mining pool. cpuminer has “-P, –protocol-dump verbose dump of protocol-level activities”, gminer has “–log_stratum arg (=0) enable/disable data of communication” and miniZ has no support for troubleshooting.

  46. Theis says:

    Hi guys,I am mining flux on single 3070 (for testing) and I would like to see ‘solved share difficulty’ parameter in the output. Could not find anything in the telemetry page. Any ideas on to get this info printed out?Kr,

  47. Miroslav says:

    Project is dead?

  48. Budi says:

    Hi miniZ team. I’m looking a way to use –mode command for my mixed gpu Rig.. I have not found a way to specify each gpu for specific mode.. Example is if gpu 0,1,2 I want to set –mode=6 then gpu 3,4,5 I want to set –mode=11.. How I set the command in my miner? For hive os.

     

    Thanks for your hardwork miniZ. You are the best ?

  49. Budi says:

    –pers auto –log –extra –templimit=65 –mode 11 –mode0 6 –mode1 6 –mode5 6 –mode6 17 –ocXreset –shares-detail –all-shares –show-solratio –show-mode –latency –coolui

    Did I set the command in the miner right for the –mode option?

    • miniZ says:

      Hi Budi,

      Thank you for your message.

      To set mode 6 to GPU 0,1,2 and to set mode 11 to GPU 3,4,5 you can use:
      --mode0 6 --mode1 6 --mode2 6 --mode3 11 --mode4 11 --mode5 11

      Because order is important you can also write:
      --mode 6 --mode3 11 --mode4 11 --mode5 11
      (This will set mode 6 to all GPUs except GPUs 3,4,5.)

      For more examples and information check our FAQ page, How to use --mode?

      If you already know the correct kernel modes, you can set the mode for each GPU and do not need to use --ocX or --ocXreset again.

      Cheers

  50. DukeNukem says:

    One of my rigs with 3070ti and 3060ti, crashes with this message: Windows successfully diagnosed a low virtual memory condition. The following programs consumed the most virtual memory: miniZ.exe (8008) consumed 13306630144 bytes, MsMpEng.exe (4168) consumed 343146496 bytes, and explorer.exe (4304) consumed 207843328 bytes.The virtual memory was set to 6x8mb 48000mb after windows was installed. Rig crashes every day at least one with the message above, pls. do you have any suggestions.

    • miniZ says:

      Hi DukeNukem,
      It seems that you may not have enough memory and need to increase the pagefile.

      Could you try to increase pagefile and try again?

      You can see here how to check the pagefile size and how to manage it.

      Let us know if this helps.

      Cheers

  51. Miroslav says:

    Could you please make and publish win zip file without the password? It makes me big problem because I can’t download it and unpack programmatically via PowerShell on my work farm. At least publish it without password on GitHub. You are the only one that enforce the password protected zip file for all the developers.

  52. ZeroII says:

    Hi miniz-Team,first thank’s of your Work. I’ve some question in my mind:I test the gtx780 and gtx960 with BTG. In specs the gtx780 have more core’s and bandwidth, so why i get-> 15sol/s with gtx960 but only-> 7 sol/s with gtx780is their any potential by tuning the algorithm for the gtx780 (Kepler)?

    • miniZ says:

      Hi ZeroII,
      Thank you for your message.
      Try adding --ocX to your command line. This will test other kernels, and may check if there are better kernels for those GPUs.
      Let us know if --ocX finds a better kernel than the default.
      Cheers

      • ZeroII says:

        Hi miniZ-Team,I got 10% more rate. The differenz between the midderange Maxwell and highend Kepler Generation, stays. I spent time to understand the Equihash-Algorithmen without going forward. But i think that the highend Kepler can profit of more parralism.If you need Tester i would be your man 😉

  53. Linkestis says:

    Hi, I have no problems, i just want to share information and ask if there is anything I can do to improve the performance. I use Windows and unfortunately the latest drivers of Nvidia 511.62.The cards are 2 x 3060 TI lhr and 2 x 3070 TI lhr.It’s an algorithm algo 125_4. For 3060TI

    CC +200,MC-0 and PL-70%. For 3070Ti – CC+200,MC-0 and PL 65%.

    miniz.exe –url=t1V2*******.****@de.flux.herominers.com:1200 –par=125,4 –pers=ZelProof –extra
    pause

    3060TI-44-45 sol/s and 3070Ti 64 -65 sol/s.

    Is there anything I can improve? Thanks for your work and I look forward to a new version ..
    Is there any hope of working with mix AMD and Nvidia in the future?

    • miniZ says:

      Hi Linkestis,
      Thank you for your message.
      In the next version the RTX30XX will have improvements for 125,4 algorithm.
      Also, AMD support will be introduced for 125,4 and for RX 570/580 GPUs.
      Cheers

      • Linkestis says:

        I look forward to the new version then.Perfect job thanks!!Regards

      • Matt S says:

        How is that new version coming along? Would love to try out a beta version with optimizations for 30 series cards…

        • miniZ says:

          Hi Matt S,
          It has been quite a challenge. However, we realize that we cannot delay the new release much longer.
          We still need to test a few modifications, but we think that it is reaching a final stage.
          Thank you for your patience.
          Cheers

          • Matt S says:

            Hi,It’s been a month since the last time we spoke, It’s also been 5 months since the last release of the miner. Maybe throw a beta out there…

  54. Jyll says:

    I tested the 125.4 algorithm on 3070 LHR. Used drivers 460.91.03 465.31 470.103.01 495.46 510.54 Marked the best.
    +190 core, pl 170W – 65 sol I’m waiting for the new version, I hope it will be even better!

    • miniZ says:

      Hi Jyll,
      Thank you for the information, and for sharing your settings!
      New new version will be a little better for the RTX on Flux mining.
      Cheers

  55. Linkestis says:

    Hi,with some time can you tell us when you will reach the final of a new version. However, many months have passed ..

  56. Thierry says:

    Bonjour, hiveos/ btc-s37/ Miniz = 7 gpu OK / 8 gpu= boot impossible sur la dernière version 1.8y4rc2 (( miniz exited (exitcode=139) wait to cooldown a bit))/ version: v1.8y2 = boot 8 gpu= ok

    • miniZ says:

      Hi Thierry.

      Are you still experiencing this issue?

      We are about to release a new version.
      We would like to ask you, if you could test your 8 GPU rig with the newer version, and then let us know if this resolved the problem.

      Thank you!
      Cheers

  57. Aurum says:

    Nvidia made the bad decision to use the {FP32 AND (FP32 OR INT)} architecture. So if integer math is used the speed is cut in half. If equihash125 mining uses integer math can it be reprogrammed to only use floating point math to speed it up?

    • miniZ says:

      Hi Aurum,

      Floating point math is usually much slower than integer math, so we think that it is unlikely that a FP implementation would do any good.

      Thank you for your message!
      Cheers

  58. dexter says:

    hi.. on flux mining is 8gpu suported?
    with 5 gpu was ok , when i add 2 more..im getting errors all the time

    ************ miniZ v1.8y4rc2 ************ Number of CUDA[8.0] devices found: 7 /root/xminer.sh: line 195: 976 Segmentation fault {minerSudo} /root/miner/${MINER_PKG_NAME}/${MINER_FILE} ${MINER_OPTIONS_GO} Miner ended or crashed. Restarting miner in 30 seconds…

    using simpleminig as OS platform

    • miniZ says:

      Hi dexter,
      Thank you for the feedback.

      We’re just about to make a release. The new release will have changes that may resolve this issue.

      Could you test it and let us know if it helps?

      Thanks a lot for the feedback.
      Cheers

  59. dexter says:

    best setting for 2080 and 2070 are +150 core-200 mem (yes, negative 200)

  60. Aurum says:

    I hope you will add Watchdog to automatically restart MiniZ. I have to keep an eye on my pool to see when computers stop mining.

    • miniZ says:

      Hi Aurum,
      You can run the miner by creating a bat file (will make the miner restart) as follows. Do not forget to customize your wallet address, worker name, server and port number:

      :loop
      miniZ.exe --url=YourWallet.worker@server:port --log --extra
      timeout /t 5
      goto loop

      Let us know if this solution works for you.
      Cheers

  61. HWND says:

    Equihash 144,5 EXCC COIN 2060S=75SOL,But they never submitted it,It will also disconnect. What is the reason

    Leave a Reply

    Your email address will not be published. Required fields are marked *

    I accept the Privacy Policy