Bitcoin Forum

Hello fellow miners. Some of you can remember me from HSRMINER Neoscrypt Fork by Justaminer (https://bitcointalk.org/index.php?topic=2765610.0). I've updated that miner, added new features and fixed bugs without access to source code.

I like Claymore's Eth miner much and used it many times in 2019/2020, so it was sad for me to see how it stopped working because it only supported up to 384 DAG epoch. I decided to see what can be done to solve the problem. After weeks of researching and hard work, the solution is ready:

---

If you are unable to run the miner after August Win10 updates, please read this (https://bitcointalk.org/index.php?topic=5305046.msg57675444#msg57675444) to solve the issue with couple of mouse clicks.

If you are using -rxboost/-strap command-line options and have problems with strap driver not starting after latest Windows updates, please read this (https://bitcointalk.org/index.php?topic=5305046.msg57480714#msg57480714) to solve the issue.

Features of the mod:

• lowest in the industry 0.6% devfee! (it means that every 95 minutes the miner will mine devfee for 35 seconds).
• command line option "-strap" now works for AMD Polaris GPUs on video drivers 20.5.1+, so now you can get speed increase over Phoenixminer! (check video from "The mod in action" section below)
• support of DAG epochs up to 600 so you can continue to use legendary Claymore ETH miner!
• fixed bugs of the original miner: it was unable to access nvml.dll on Windows 10, so temperature monitoring, fan control and power draw stats were unavailable.
• added support of new Nvidia drivers 460.89+, miner initializes NVML properly with these drivers, so temperature monitoring, fan control and power draw stats are available.
• now miner shows AMD GPUs power draw stats on newer AMD drivers (tested with 20.12.1)
• now miner shows AMD driver version at startup and warns if driver's version is not supported (i.e. lower than 20.5.1)
• now miner saves AMD Polaris GPUs' core and memory clocks/voltages at startup (before applying custom ones with -cclock,-mclock,-cvddc,-mvdcc) and restores it during exit.
• now miner is able to apply -cclock values to Navi GPUs on new video drivers.
• now miner saves AMD Navi GPUs' core clocks/voltages and memory clocks at startup (before applying custom ones with -cclock,-mclock,-cvddc) and restores it during exit.

Currently this mod supports only Windows version of the miner, Nvidia GPUs with at least 6 GB of video memory, and several AMD GPUs with at least 8 GB of video memory: Polaris RX 470,480,570,580,590, Navi RX 5500XT, 5600, 5600XT, 5700, 5700XT, Vega. Please read "AMD GPUs support" section below for details.


~ Increased maximum number of supported DAG epochs to 600.

---

You can see that my mod works on this video: https://vimeo.com/495752694 (https://vimeo.com/495752694)

If you have 8GB AMD Polaris GPU (RX470/480/570/580/590), you can get speed increase over Phoenixminer (+0.6 MH/s per gpu) by using -strap and -rxboost options, example of hashrate you can get for RX580 with my mod https://vimeo.com/533071364 (https://vimeo.com/533071364) vs Phoenixminer https://vimeo.com/533072304 (https://vimeo.com/533072304)

See full details and screenshots in this msg (https://bitcointalk.org/index.php?topic=5305046.msg56715263#msg56715263).

---

Warning!  MEGA.NZ closed many accounts with miners, including my one. I've reuploaded my mod to github:

https://github.com/just-a-miner/moreepochs/releases (https://github.com/just-a-miner/moreepochs/releases)

Checksum for MoreEpochs_Mod_of_Claymore_ETH_Miner_v15Win_by_JustAMiner_v2.7.zip
SHA256: 97cc16d88282d05f7c086d8c4d384750bd9163bc1c2e411cec041b19e1f318e1

Checksum for MoreEpochs_Mod_of_Claymore_ETH_Miner_v15Win_by_JustAMiner_v2.6.zip
SHA256: f2006f4803566bd11cea99cf731844ae7e499d2154db8b3c015ecc76b226e6f0

Checksum for MoreEpochs_Mod_of_Claymore_ETH_Miner_v15Win_by_JustAMiner_v2.5.zip
SHA256: aa45e98b92a8299da9f5e0dede7cb1ad42a0cc88520ca18404619ed63e9e5bae

Checksum for MoreEpochs_Mod_for_Claymore_ETH_Miner_v15Win_by_JustAMiner_v2.4.zip
SHA256: 6ca9487932702f0c6911a0c45bca21d7babd70ecef479f79e6113a646e042e3a

Checksum for MoreEpochs_Mod_for_Claymore_ETH_Miner_v15Win_by_JustAMiner_v2.3.zip
SHA256: 06cf9c4cddde88fda8011989ee868b763c4be00306c9b6925ae758e679a06f2a

Checksum for MoreEpochs_Mod_for_Claymore_ETH_Miner_v15Win_by_JustAMiner_v2.2.zip
SHA256: 631a8152fbe9ecf1fd6b0ad0fcde93b8e1822a871ba05894720b6b3d5af7f109

Checksum for MoreEpochs_Mod_for_Claymore_ETH_Miner_v15Win_by_JustAMiner_v2.1.zip
SHA256: 1dac907f65aaca8088077c665bc690fecf4d8ab0a25989ae4383d8086642b8bc

Checksum for MoreEpochs_Mod_for_Claymore_ETH_Miner_v15Win_by_JustAMiner_v2.0.zip
SHA256: c0a852529fd9dc6a36b728d89bd4b73e91751debd85f7e2a785cd23efc4c30fb

Checksum for MoreEpochs_Mod_for_Claymore_ETH_Miner_v15Win_by_JustAMiner_v1.9.zip
SHA256: 5523631e49f09160499a1ecdcb861f4633fd935c5e76f22e925a573a52b36a6d

---

Unpack the archive and run the miner as you usually do. If the mod will complete its job successfully, you will see green colored message below Claymore's logo:

https://i.ibb.co/gS2wvKj/More-Epochs-How-it-looks.png

and now the miner is able to work with DAG epochs up to 600.

If you don't see this message, make sure you are using my original mod from my download link. If it still doesn't work for you, contact me and I will  try  to help you.

---

Video drivers:

You should use AMD video driver 20.5.1 or newer (most tests were done on driver 20.12.1), because older drivers can't allocate 4+ GB video buffer
in single chunk and that will prevent the miner from creating 4+ GB DAG.


Issues:

- 8 GB R9 390 doesn't work.

This issue can be fixed in next versions of the mod.


List of working GPUs:

AMD GPUs that work fine (tested on AMD video driver 20.12.1)

- 8 GB AMD Polaris: RX 470, 480, 570, 580, 590
- 8 GB AMD Navi: RX 5500XT, 5600, 5600XT, 5700, 5700XT
- 8+ GB AMD Vega GPUs


"-strap" and "-rxboost" options:

Command line options "-rxboost" and "-strap" (since v2.0 of the mod) works with AMD Polaris GPUs on video drivers 20.5.1+ (tested on AMD video
drivers 20.5.1 and 20.12.1). These options apply fast memory timings and greatly increase mining speed.
If you are going to use these options first time, you will need to do next steps once:

1) run miner as administrator once with "-rxboost 1" option. Miner should automatically uninstall old strap driver and install new one. If miner was unable to do auto uninstall/install of new strap driver, see step 2) and 3).

2) in miner folder go to "\strap_driver" subfolder and run as  admin "uninstall_strap_driver.cmd". This will stop and uninstall old strap
driver if it was installed.

3) run miner as admin with "-driver install" option.

If miner successfully installed new strap driver, you will see next message:


After that you don't need to run miner as administrator anymore and you can use -rxboost and -strap options.

Note that rxboost/straps now apply after DAG generation to avoid possible DAG corruption. When DAG generation is finished and if rxboost/strap applied
successfully, you will see next messages :


where x is GPU index that rxboost/strap applied to and "..." is specific strap.


Please be aware that some software can stop straps from working. For example, monitoring tool HwInfo does that, if you started that software then straps won't have effect on mining speed, and you have to completly reboot rig to make them work.


Please note that both "-rxboost" and "-strap" can crash/freeze your rig because not every GPU is able to work with faster timings especially at high memory
clocks. Refer to "Readme!!!.txt" to get more information about how to use -strap and -rxboost.


Compute Mode:

If after installing new AMD driver your AMD GPUs hash at very low speed, e.g. RX 580 shows like 9 MH/s instead of 27-33 MH/s, you need to enable Compute Mode in AMD Drivers. To do this, you need to run the miner as administrator once and after you see message:


press y key. After that you should see next message:


Exit the miner and reboot your rig. Now you can run miner as usual and your
AMD GPUs should work at full speed.

---

By default, there is Cuda 8 vesion of file EthDcrMiner64.exe in the root folder of the miner. If you want to use Cuda 10 version of Claymore ETH Miner  (it can give some speed increase), copy cudart64_100.dll and  EthDcrMiner64.exe from the cuda 10 subfolder to the root folder of the miner and overwrite EthDcrMiner64.exe there.

Then you can run the miner as usual and you should see:

 CUDA Driver Version/Runtime Version: xx.x/10.0

As you can see, Runtime Version is 10.0. It means miner uses Cuda 10.0 API.

If you want to get back to Cuda 8, unpack file 'EthDcrMiner64.exe' from the  mod's archive root folder, and copy it to the root folder of the miner, overwriting 'EthDcrMiner64.exe' there.

After that if you'll run the miner you should see:

 CUDA Driver Version/Runtime Version: xx.x/8.0

As you can see, Runtime Version is 8.0. It means miner uses Cuda 8.0 API.

Cuda 8 can be slightly faster on some rigs, and Cuda 10 can be slightly faster on others. You can test both versions on your rig and choose fastest one.

---

     Version: 2.7

~ Increased maximum number of supported DAG epochs to 600.

     Version: 2.6

~ Added unsigned version of strap driver that can work with Windows Update KB5004237. Unsigned driver requires Windows Test Mode.
  To enable Windows Test Mode make sure that "Secure Boot" is disabled in UEFI BIOS, then run "Enable_Windows_TestMode.cmd" from
  "strap_driver" folder in miner's archive as admin and reboot rig.
  To install unsigned strap driver run "install_Unsigned_strap_driver.cmd" (don't run as admin, it will ask for elevation after run).
  You can also install unsigned strap driver manually by running miner as admin with "-driver install_test" option.

     Version: 2.5

~ Devfee is reduced to lowest in the industry 0.6% ! (it means that every 95 minutes the miner will mine devfee for 35 seconds).

~ Now miner applies straps only once after first DAG generation at start, and doesn't reapply straps again in case of epoch's change during mining to
  increase stability.

     Version: 2.4

~ Improved straps apply method for Polaris GPUs. Most issues when miner reports that strap was applied successfully but hash speed doesn't increase are solved. Custom memory timing level that was set in AMD Software (Wattman) doesn't affect straps apply success anymore.

~ Now miner warns if Compute Mode is not enabled for AMD GPUs.

     Version: 2.3

~ Now miner saves AMD Navi GPUs' core clocks/voltages and memory clocks at startup (before applying custom ones with -cclock,-mclock,-cvddc)
  and restores it during exit.

~ Fixed bug that can occur during save/restore of clocks/voltages for AMD Polaris GPUs.

     Version: 2.2

~ Now miner saves AMD Polaris GPUs' core and memory clocks/voltages at startup (before applying custom ones with -cclock,-mclock,-cvddc,-mvdcc)
  and restores it during exit.

~ Now miner shows AMD driver version at startup and warns if it's not supported (i.e. lower than 20.5.1).

~ Fixed bug of original miner with applying -cclock values to Navi GPUs on new video drivers.

~ Several minor fixes.

     Version: 2.1

~ Changed behavior of "-driver uninstall" option. By default, miner also tries to disable testsigning mode during strap driver uninstall. This
  often leads to error on systems with SecureBoot enabled and prevents miner from uninstalling strap driver. Now when miner started with
  "-driver uninstall" option, it only uninstalls strap driver without trying to disable testsigning mode.

~ Command line option "-mode" now has default value of 1, which means that if this option is not specified, dual-mining mode is disabled by default
  and miner works in ETH-only mining mode. This is usefull for AMD GPUs (Polaris), because allow miner to get best mining speed by either automatic or
  manual fine-tuning of -dcri values.

~ Added messages that straps/rxboost will be applied after DAG generation if strap/rxboost options used. Remade several messages that miner
  displays in case of error during installing/uninstalling strap driver to give user more clear instructions.

~ Several small fixes.

     Version: 2.0

~ Command line option "-strap" now works for AMD Polaris GPUs with video drivers 20.5.1+ (most tests were done on 20.5.1 and 20.12.1).
    Read "AMD GPUs support" section for details.

     Version: 1.9

~  Fixed issue with incorrect shares on 8+ GB Vega AMD GPUs.

     Version: 1.8

~ Fixed issue with incorrect shares on 8 GB Polaris AMD GPUs: RX 470, 480, 570, 580 (tested on AMD video driver 20.12.1).
~ Command line option "-rxboost 1" works fine with Polaris AMD GPUs (tested on AMD video driver 20.12.1). For now avoid using "-strap" command line
   option for AMD GPUs, it will crash rig. This problem will be fixed in next versions of the mod.

     Version: 1.7

~ now miner shows AMD GPUs power draw stats on newer AMD drivers (tested with 20.12.1)

     Version: 1.6

~ Improved devfee connection handling. Reduced retry delay to 1 second between connection attempts to other devfee pools if miner can't connect to default devfee pool when devfee starts.
    Reduced retry delay to 1 second between connection attempts to other devfee pools if miner lost connection to devfee pool during mining devfee.

~ several minor bug fixes and improvements

     Version: 1.5

~ Fixed issue with incorrect shares on Navi AMD GPUs. Other AMD GPUs will be addressed in next versions of the mod.

     Version: 1.4

~ Added support for Nvidia drivers 460.89+, now the miner initializes NVML properly with these drivers, so temperature
   monitoring, fan control and power draw stats are available.

     Version: 1.3

~ Fixed Claymore's bug that lead to error "NVIDIA NVML library not found, temperature monitoring for NVIDIA GPUs disabled" on  Windows 10.
   Nvidia driver 456.71 is tested on Windows 10 and shows temperature, fan speed, power draw of Nvidia GPUSs, and allow to control GPUs fans.
   Nvidia driver 460.89 still gives the error "NVML API: Init Error 1" - this will be fixed in next releases of the mod.
  
     Version: 1.2

~ Several AMD GPUs are now enabled to work with DAG epochs 385+ :
   R470/R480, Vega, Navi. This is test support, please report any problems.

     Version: 1.1

~ added support of Cuda 10 EthDcrMiner64.exe

~ fixed bug with buffer allocation for DAG epochs 388+

     Version: 1.0

~ First version, supports only Cuda 8 EthDcrMiner64.exe and Nvidia GPUs.

---

Several people asked me about devfee in this mod, and I want to explain what is my reward for creating and supporting it. In this mod I've done huge amount of unique work from the start and even more work afterwards as I continue to suport and improve it for several months, so people can continue to use legendary Claymore miner. As a reward for that hard work which as I hope will help many people, I pointed devfee to my eth wallet. This will support me and let me to invest more time into development of future versions of this mod. I think this is fair, because for some reason original developer disappeared, his miner stopped working and so devfee mining to his wallets stopped too anyway. Size of devfee is reduced, it's now lowest in the industry 0.6% ! It means that every 95 minutes the miner will mine devfee for 35 seconds.

---

Feature versions of this mod will have support of Linux version of the miner.


Enjoy!

I know it was a little late release, but I did my best.

I checked it in a rig with 1080ti's. Works. Thank you) But the t-rex miner 19.4 cuda 9.1 gives me same MH from the card. The original claymore 15 gave more. It probably used cuda 9

thanks for hard work please AMD very important, add amd support under windows you king man

Yes Add AMD support ;)

Hi Just_a_miner - I know with the HSRMINER that you'd also made it devfee free - is the devfee portion unchanged on this version of claymore, gone, or pointed somewhere else?

Seems to be working fine with my nvidia gpu's after just a few minutes. Similar to others I'd love to run it with AMD gpu's as well. Thanks!

Thanks for the feedback. There are two Cuda versions in Claymore's archive. Cuda 8 version is in root folder (EthDcrMiner64.exe and cudart64_80.dll) and Cuda 10 version is in 'cuda 10' subfolder (EthDcrMiner64.exe and cudart64_100.dll). To change Cuda version to 10 you had manually copy .exe and .dll from 'cuda 10' subfolder to root folder and overwrite Cuda 8's .exe and .dll there. I removed 'cuda 10' subfolder from the archive because my mod doesn't support it yet.

So, may be you used cuda 10 version of EthDcrMiner64.exe before so it gave you more Mhs. I will add support of Cuda 10 EthDcrMiner64.exe in next version of my mod.

Why would anyone use this sketchy stuff when there are perfectly working miners out there?

Thank you for your kind words  :) I will try to do that.

HSRMINER Fork was devfee free only at the beggining of development. After many improvements I've added devfee to the HSRMINER Fork.
In this mod I've done huge amount of unique work from the start, so people can use this legendary miner again. As a reward for that hard work which as I hope will help many people, I pointed devfee to my eth wallet. I think this is fair, because for some reason original developer disappeared, his miner stopped working and so devfee mining to his wallets stopped too anyway.


Thank you, I will try to add AMD support.

I'm not agree with the "sketchy stuff" part, but I understand that people can be suspicious. As for the reasons for using my mod, there are plenty, but I think main reason is that this is legendary miner people loved to use, and with my mod they have unique opportunity to continue to use it.

Hi, can you support Hawaii AMD Radeon R9 290 4Gb cards ?

dont pay attantion, there will be always comment in every types. you did right thing nothing wrong with it, keep great working. looking forward for devolpments, happy new year

Is there a devfee for you?

Would be nice if you explain little bit what this mod does?
If no devfee, why not open source?

19:28:21:717   fd8   ETH: 12/31/20-19:28:21 - New job from asia.ethash-hub.miningpoolhub.com:20535
19:28:21:718   fd8   target: 0x000000007fff7ffe (diff: 8590MH), epoch 384(4.00GB)
19:28:21:743   fd8   ETH - Total Speed: 20.107 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
19:28:21:749   fd8   ETH: GPU0 2.515 Mh/s, GPU1 2.516 Mh/s, GPU2 2.511 Mh/s, GPU3 2.515 Mh/s, GPU4 2.515 Mh/s, GPU5 2.515 Mh/s, GPU6 2.515 Mh/s, GPU7 2.506 Mh/s
19:28:33:390   1590   Quit, please wait...
Why am I only 2mh? What did I do wrong

what cards are they?

570 8g power colo vram samsung

devoloper said it is not ready for AMD yet. am waiting too for AMD release,

ok thanks ;D

Hello, yes there is devfee for me:


For now mod adds support of epochs 385+ for Claymore ETH miner. But only Cuda 8 version of Claymore miner supported, only Nvidia gpus and only Windows version. In future versions of my mod I'm going to add support of Cuda10 for Nvidia gpus, support for Linux version and support of AMD gpus.

Yes, hopefully in a week  :)

thank you for your kind words and support, happy new year

I look forward to, thanks

And the size of the devfee? It helps you to gain the trust if you are open about these things.

Accidentally posted on a thread bumping account?  ::)


EDIT: LOL he deleted his post.  It says a lot about someone that is making fresh accounts to post on his main accounts topics.  I would avoid this as its potential malware.

Size of devfee is unchanged, it's 1% - 36 seconds of every hour.

You even created new account to post copy of my message, made quote from it and deleted message, then you accused me of bumping the thread from that account. What a dedication. Are you on payroll?  :D

Dude, I got it - you don't like my mod. You didn't like Phoenixminer either and made it clear in their topic too:

https://bitcointalk.org/index.php?topic=2647654.msg27082971#msg27082971 (https://bitcointalk.org/index.php?topic=2647654.msg27082971#msg27082971)
https://bitcointalk.org/index.php?topic=2647654.msg27086637#msg27086637 (https://bitcointalk.org/index.php?topic=2647654.msg27086637#msg27086637)
https://bitcointalk.org/index.php?topic=2647654.msg27118193#msg27118193 (https://bitcointalk.org/index.php?topic=2647654.msg27118193#msg27118193)
https://bitcointalk.org/index.php?topic=2647654.msg27148871#msg27148871 (https://bitcointalk.org/index.php?topic=2647654.msg27148871#msg27148871)

Now when you said that you don't trust my mod for two times already, it's time for you to stop. Because if you post your personal opinion and concern, two times are more than enough, I heard you and you are free to not use my mod. Of course if you are on payroll you won't stop, will you?  ;)

Come on guys nobody cares about devfee, i will pay 1% as long as this miner works for me. just make it work good old days apply straps was amazing good speeds on amd.

Actually, Claymore abbandoned further miner developement and he abbandoned us all when we needed him the most (4+ GB DAG) so it really would not be fair if devfee would still go to him...

If the devfee is 100% it's quite big problem, thats why i asked but 1% np.

But @just_a_miner thanks for answering the questions.

A lower fee would be fair.
You didn't develop Claymore from scratches.
Dropping it to 0.5% would be a nice move toward the community.

New Version just released!


Changelog:

~ added support of Cuda 10 EthDcrMiner64.exe

~ fixed bug with buffer allocation for DAG epochs 388+


Update to avoid error with epoch 388+ !

Also first message of the topic (https://bitcointalk.org/index.php?topic=5305046.msg55975429#msg55975429) updated with info about how to switch between Cuda 8/10 versions of the miner.

Thanks for update, still waiting for AMD support. Especially polaris cards.

I was thinking about reducing devfee to "modern standarts" of 0.65%-0.7%. I just don't wanna touch this part of the software right now, because I'm quite busy with adding support of Linux version and AMD Gpus. Devfee is probably protected well, so I can't change it in a hurry. After adding support of Linux version and AMD Gpus, I will take a closer look at devfee part of the software and probably will be able to reduce it.

Everything is according to the roadmap - I'll get amd gpu in a 1-2 days, and will start to add support to the mod  :)  Can you share what amd gpu do you have?

I think you should focus on Polaris (RX470/570/570/580) and Navi (RX5600/RX5700) cards

Sounds great
Mostly rx 470,480,570,580
then rx 5700

support for 3080?

thank you so much for reviving the miner

No work, no text @* More Epochs v1.0 mod by JustAMiner [ justaminer@tutanota.com ] *@ chek sha256 right

Understand, I guess you are working on the assembler level ?

dont work new releases

HAWAII please! R9 390 8gb .... thanks and keep the good vibes

Can confirm 1.1 not loading the plugin.

---

https://mega.nz/folder/dNlw2Lzb#7hVfm7oADQV2wMLqVi4erQ (https://mega.nz/folder/dNlw2Lzb#7hVfm7oADQV2wMLqVi4erQ)

---

Ok, I see.

right, most cards on the market I guess rx 580. would appreciate if you could support the rx 580 :)

Many thanks. Operation "Watch the RTX 3700 recoup its cost via the Awesome Miner UI for shits and giggles" is back on due to your kindness. =)

Nice, thanks  :)

AMD for, how are the situations?
Thank you for your work

Thank you.

Currently I add support for AMD Gpus. Hopefully new version of the mod will be released in 2-3 days.

We can talk about adding new features and gpus support after I'll make AMD Gpus work with epochs 385+ and when Linux version of the miner will be able to works with epochs 385+.

Trying it on 4 1070 rig
Works, but it says that Nvidia SMI is not found thus no temperature control. any idea?
Works in Cuda 10 also

Also, i can't see the temperature of the cards and how many watts they draw, is it a known issue?

What driver you use? I think you driver is a newer one, and claymore 15.0 is very old, so maybe it will not support your driver fully.

Seems like new Nvidia driver problem, they've probably changed something. I also tested my mod with 1070, but with some older driver like 441.xx and it for sure shows temperature, fan speed and power draw. I'm not updating driver because 1070 is quite old. Nvidia won't improve their drivers for mining, all they do is add support for new games and break something along the way  :) So there is no point in using new drivers with older GPUs for mining.

You can try to download older driver from here: https://www.geforce.com/drivers/beta-legacy (https://www.geforce.com/drivers/beta-legacy)

Later I will fix the issue with newer drivers.

Can you tell what is your nvidia driver version?

I have 432.00 ...
it says:
Maybe I should install your version 444 and try...

Latest version 461.09
my card 3080

This seems doesn't work on my AMD RX580 8gb. Any solution to this?

https://bitcointalk.org/index.php?topic=5305046.0


read topics :/

Im waiting for AMD too. I have rigs wit 570 8Gb.

I will take a look what can be done about that. Do you use Win10?

Does miner say "NVIDIA NVML library not found, temperature monitoring for NVIDIA GPUs disabled" at start?

New Version just released!


Changelog:

~ Several AMD GPUs are now enabled to work with DAG epochs 385+ :  R470/R480, Vega, Navi. This is test support, please report any problems.
   If you only have NVIDIA GPUs,  you can stay on v1.1


 

Works Perfectly Great Mining Ethereum on Radeon Rx 580 8GB.
Thank you so much  :)

I test on 6x RX 570 8Gb, mixed with Samsung and Hynux memory.
Miner works like good old Claymore but i recived REJECTED SHARES.
Evry three to five share was rejected (cpu 1150, mem 2150, boath 920 mV, 182mh/s).
I lower hashrate to very stable (cpu 1100, mem 2100, boath 920 mV, 175mh/s) and I gor rejected shares to.

What AMD driver version are people using? I'm using a fairly old one probably because I wanted one that works with OverdriveNTool and had noticed some of the newest ones don't. I'm using version 20.4.1 and with this driver and MoreEpochs v1.2 none of my AMD GPU's are able to create their DAG files.

Thank you for the report.


Thanks for the information, I will investigate the issue.


I was testing with driver version 20.12.1 What AMD GPUs do you have? What kind of error did you get when gpu wasn't able to create DAG?

New Version just released!


Changelog:

~ Fixed Claymore's bug that lead to error "NVIDIA NVML library not found, temperature monitoring for NVIDIA GPUs disabled." on
  Windows 10. Nvidia driver 456.71 is tested on Windows 10 and shows temperature, fan speed, power draw of Nvidia GPUSs, and
  allows to control GPUs fans. Nvidia driver 460.89 still gives the error "NVML API: Init Error 1" - this will be fixed in
  next releases of the mod.

If you have AMD GPUs only, you can stay on v1.2

I was testing with driver version 20.12.1 What AMD GPUs do you have? What kind of error did you get when gpu wasn't able to create DAG?

I use Adrenalin version 20.8.1.

miner star fine with my "Hawaii", hashing at 15.4mh/s in ethermine.org:5555(ssl) using all the server pool locations, but after some minutes realize that none share was submitted , then , get this red messsage "GOT INCORRECT SHARE, IF YOU SEE THIS WARNING OFTEN, MAKE SURE YOU DID NOT OVERCLOCK IT TOO MUCH!" ... am using Adrenalin-2020-20.11.2 with my Gigabyte R9 390 8gb

Awsome upgrade. thank you really appreciated.i wish you keep ugrading this software.
I forgot to choose gpus in claymore, does anyone remember how to choose? Example in phoenix -gpus 1,2,5,6,7,8,9
Thanks, happy mining
edit: nevermind founf it its -di

I can confirm that 1.3 works on Nvidia and nvml is fixed

Rig 1 has a Vega 56, and an RX580; rig 2 has 3x RX580's - all are 8GB. I get the same error on all of them when I try to run version 1.2

Setting DAG epoch #387 for GPU3
Create GPU buffer for GPU3
GPU3 - not enough GPU memory to place DAG, you cannot mine this coin with this GPU
GPU3 - OpenCL error -61 - cannot allocate big buffer for DAG. Check readme.txt for possible solutions.

I'm guessing the driver version is too old and I'll need to go with something slightly newer, but I'm going to wait until I can get onsite to use DDU driver uninstaller and then update.

Anyone have OverdriveNTool running with a newer driver version - if so, which version? Thanks.

try newer drivers for me its working. just tested on my desktop soon will switch rigs.thanks we are back again  8)

Need help here. system win10 has 9gpus rx 480 rx470. if i use strap command rig crashes. without strap working just fine. Does anybody faced this problem? thanks
edit : it looks like driver issue :( i have 20.11.1.
  @just_miner could you please check this strap command. i remember it works with blochain driver and old driver back in 2019. really dont wanna go back with that drivers again, will super if strap command runs under 20.11.1 driver or newer version. thanks

R9 390 is not supported now, it should not work.


I will take a look at it later, won't promise it will be solved though.

Thank you, but I repeat, currently 390 is not supported, so it won't find shares.

New Version just released!


Changelog:

~ Added support for Nvidia drivers 460.89+, now the miner initializes NVML properly with these drivers,
   so temperature monitoring, fan control and power draw stats are available.

If you have AMD GPUs only, you can stay on v1.2

hello, when I try different programs, I can use it without any problems, but I get an "GPU got incorrect share. If you see this warning often, make sure you did not overclock it too much!" error in the claymore mod program. Is there anything I can do?

Comes every 10 to 15 seconds on almost all cards, not on a single card. The same error comes across even though I try to reduce the memory and clock values by 50%.

The version I am using AMD 20.12.1
Card: 8G Sapphire Rx 580 Nitro + and 8G Sapphire Rx 580 Pulse.

for example when i run phoenixminer on the same computer i don't get any errors.

I'll be glad if you can help, thanks.

Thanks for the information, I know about the issue. Currently support of AMD Gpus is in test state in this mod, when I'll solve the issue I'll release new version of the mod.

One user sent me .log with similar error:

---

GPU #0: Ellesmere (Radeon RX 580 Series), 8192 MB available, 36 compute units (pci bus 1:0:0)
gpu 0 max alloc size is 4048 MB
Setting DAG epoch #388(4.03GB)...
Setting DAG epoch #388 for GPU0
Create GPU buffer for GPU0
GPU0 - not enough GPU memory to place DAG, you cannot mine this coin with this GPU
GPU0 - OpenCL error -61 - cannot allocate big buffer for DAG. Check readme.txt for possible solutions.

---

As you can see GPU reports 8192 MB of vram, but other api call returned value of 4048 MB which is not enough for 4.03 GB Dag.
The situation probably happens with specific models of GPUs or with certain version of drivers.

Can you check your .log file to see if there is same problem? Also what is your OS (Win7/Win10) and AMD driver version?

I checked my log and it is the exact same. Do you know what driver that person was using? I'm using version 20.4.1 on Win 10. It will likely be another week or two and I should be able to try loading new drivers.

So just trying this out on my 8GB Sapphire RX580 and it tells me it can't load the DAG because it's too large. This is odd because it's an 8GB card. Is there some parameter I am missing? Well I just re-read the post above and since I have same type of card, that might be the same issue..

new epoch +389 change allocation in config claymore and other miner change "-eres 2 to -eres 1 or 0"

I tried that and same thing. I just went back to Phoenix miner since I was able to get that to work under mOS instead of Windows.

No, I don't know his driver version. But you told me your driver version, so I will check if this version will give me wrong vram size too.



Yes, probably same issue. If I'll find the workaround I'll inform about it here.

The problem is in AMD drivers, you can read more here https://github.com/ethereum-mining/ethminer/issues/1966 (https://github.com/ethereum-mining/ethminer/issues/1966), the solution is here https://github.com/ethereum-mining/ethminer/issues/1966#issuecomment-663826059 (https://github.com/ethereum-mining/ethminer/issues/1966#issuecomment-663826059)

So you need newer AMD driver, 20.7.2 or later. Then GPU will be able to allocate more than 4gb of vram in single chunk.

hi, hope your fine.
ok, so the question is... have you the plan to make support for R9 390 8GB (HAWAII) or not? .. .thanks

I upgraded to AMD driver 20.7.2 and while I now can get it to run with my RX580's and Vega56, I'm having the exact same problem as batsonxl - if I have straps set to 1 or whichever straps I'd previously used, the whole rig crashes the moment it tries to applies the straps. If I set it to 0, it will start up and run, but probably with around a 35% reduction in hashrate as the straps aren't being applied. Hopefully straps can get working again for AMD's? Thanks

Hello, thanks I'm fine, just busy investigating the issue with rejected shares for AMD gpus. If I'll solve the problem, I will try to add support for 390.

With -straps 0 do you have accepted shares? Do you have shares that didn't pass check on CPU? ("got incorrect share" message in log file)

About crashing with -straps 1. The miner uses heavily protected custom kernel-mode driver to apply straps and it looks like it's not compatible with new amd gpu drivers. I think the only realistic solution for me is to use some 3rd party driver to apply straps instead of default miner's driver. Still this will require a lot of researching and therefore time.

With -straps 0 yes I get accepted shares, but I would say about 1/2 -2/3 are "incorrect share". I've also tried stock clock settings instead of overclock since my OC's were set using the original straps, but even at stock clocks I still get mostly "incorrect shares" so my effective hash is probably 1/3 of what it used to be. Is this something you think you'll be able to get working, or should I look at using a different miner for my AMD's?

Thanks for your hard work,
If claymore will NOT work with straps then there is no point to use it. i used clay only because of this, with 3rd party software all miner works fast, the point is claymore was single miner with all settings you need, make bat file thats it.you dont need oc software or mod the bios.

I see, I'm working on solution to get rid of incorrect shares for AMD Gpus. If you have many incorrect shares for your AMD Gpu and some other miner works well for you,  of course you should use it. When I'll solve the problem I will release new version of my mod (probably as separate release for AMD Gpus because Nvidia Gpus work fine already) and you will be able to test it and decide what is best for you.

Thanks. That makes sense, let's hope problem will be solved.

thanks, ill be waiting anxiously.. .good vibes!

Should work fine.

hi
in Iran dev fee doesn't work properly

all miner work good except this miner

DevFee: ETH: Stratum - connecting to 'eu1.ethermine.org' <172.65.207.106> port 5555 (SSL/TLS)
ETH: 02/03/21-14:28:52 - New job from eu1.ethermine.org:4444
ETH - Total Speed: 137.800 Mh/s, Total Shares: 158, Rejected: 0, Time: 01:20
ETH: GPU0 29.793 Mh/s, GPU1 38.815 Mh/s, GPU2 38.772 Mh/s, GPU3 30.420 Mh/s
ETH: 02/03/21-14:28:56 - New job from eu1.ethermine.org:4444
ETH - Total Speed: 137.858 Mh/s, Total Shares: 158, Rejected: 0, Time: 01:20
ETH: GPU0 29.793 Mh/s, GPU1 38.836 Mh/s, GPU2 38.828 Mh/s, GPU3 30.401 Mh/s
ETH: 02/03/21-14:28:59 - New job from eu1.ethermine.org:4444
ETH - Total Speed: 137.748 Mh/s, Total Shares: 158, Rejected: 0, Time: 01:20
ETH: GPU0 29.806 Mh/s, GPU1 38.773 Mh/s, GPU2 38.751 Mh/s, GPU3 30.419 Mh/s
ETH: 02/03/21-14:28:59 - New job from eu1.ethermine.org:4444
ETH - Total Speed: 137.748 Mh/s, Total Shares: 158, Rejected: 0, Time: 01:20
ETH: GPU0 29.806 Mh/s, GPU1 38.773 Mh/s, GPU2 38.751 Mh/s, GPU3 30.419 Mh/s
ETH: 02/03/21-14:29:03 - New job from eu1.ethermine.org:4444
ETH - Total Speed: 137.870 Mh/s, Total Shares: 158, Rejected: 0, Time: 01:20
ETH: GPU0 29.783 Mh/s, GPU1 38.834 Mh/s, GPU2 38.818 Mh/s, GPU3 30.435 Mh/s
ETH: 02/03/21-14:29:07 - New job from eu1.ethermine.org:4444
ETH - Total Speed: 137.793 Mh/s, Total Shares: 158, Rejected: 0, Time: 01:20
ETH: GPU0 29.793 Mh/s, GPU1 38.794 Mh/s, GPU2 38.772 Mh/s, GPU3 30.434 Mh/s
ETH: 02/03/21-14:29:09 - New job from eu1.ethermine.org:4444
ETH - Total Speed: 137.643 Mh/s, Total Shares: 158, Rejected: 0, Time: 01:20
ETH: GPU0 29.784 Mh/s, GPU1 38.764 Mh/s, GPU2 38.726 Mh/s, GPU3 30.370 Mh/s
ETH: 02/03/21-14:29:09 - New job from eu1.ethermine.org:4444
ETH - Total Speed: 137.643 Mh/s, Total Shares: 158, Rejected: 0, Time: 01:20
ETH: GPU0 29.784 Mh/s, GPU1 38.764 Mh/s, GPU2 38.726 Mh/s, GPU3 30.370 Mh/s
ETH: Stratum - Cannot connect to eu1.ethermine.org:5555
DevFee: ETH: Stratum - Failed to connect, retry in 20 sec...

hi, probably ssl port 5555 is closed by your internet provider. If you are mining to ethermine, original miner would try to use ssl connection to mine
devfee there too. I can change this in next versions of the mod if you won't be able to solve the problem with port 5555 on your side.

New Version just released!


Changelog:

~ Fixed issue with incorrect shares on Navi AMD GPUs. Other AMD GPUs will be addressed in next versions of the mod.

If you don't use AMD Navi GPUs, you can stay on older versions of the mod.

any indicative hashrate with 4 gb cards  rx 470/570?

4 gb GPUs aren't supported due to lack of video memory for holding 4+ gb DAG for current and future epochs.

did you solve issue with closed 5555 port? I'll release new version soon with workaround.

If you pair them in a system with a 8GB 470/570 then they will mine but at a reduced hashrate. Best I could get on my 470 4gb cards was 11 MH/s using 1.5

anybody here can tell me with the R9 390 8gb WHAT IS THE actual AVERAGE hashrate for ETH? ...using what miner and overclocked or not? ill appreciate that info!!! thanks

New Version just released!


Changelog:

~ Improved devfee connection handling. Reduced retry delay to 1 second between connection attempts to other devfee pools if miner
   can't connect to default devfee pool when devfee starts. Reduced retry delay to 1 second between connection attempts to other
   devfee pools if miner lost connection to devfee pool during mining devfee.

~ several minor bug fixes and improvements

Can you be more specific please. So you are using v1.5 of my mod with several 8gb AMD Gpus and several 4gb AMD Gpu in same rig. And 4gb Gpus are able to mine, but with reduced hashrate. Are you sure that 4gb Gpus actually found shares?

Hello, Will you have a work in Linux?

Hello, yes. But sadly I can't tell when. Right now I have next roadmap:

1) fix issue with incorrect shares for 8gb AMD Polaris Gpus (470/570/480/580) and Vega.

2) fix crash when applying straps for AMD Gpus.

3) Linux version support.

Current progress:

1) is 85% done, but I need Polaris GPU to test, friend will send me his 570 or 580 to test in 3-4 days.

2) can take a lot of time, will see.

After that I will start adding support for Linux version of the miner. If I'll find more time, I'll start working with Linux version in parallel with other tasks.

LOLMINER gives nearly full hash 27-28mhs for 4gb on ETH right now. 
Does full tuning by card.

Or you can stay using outdated software with stale share problems and lacking performance.  I get better hash now with LOL/TRM/PM then claymore.  Let it die man.  Your getting what 12mhs or less on 4gb cards.  And 8gb cards you getting stales and crashing on AMD. 

If you are a good dev then step up and write a full miner...if not slink back to the shadows and quit ridding claymores dead software and his coattails.

4g 27-28mhs ???? how?

Not working whit older cards like claymore made it work.



3.5 compute is not compatible anymore no point using it if it dont support old cards.


thanks anyways for your try.

6+ GB Nvidia GPUs work fine.  8 GB AMD Navi works fine. Polaris 8 GB GPUs will work fine after I'll fix what have to be fixed for supporting of 4+ GB DAG.

4 GB AMD GPUs are not supported. Anyone with 4 GB AMD GPUs should use some other miner.

Thanks, what older cards do you mean?

New Version just released!


Changelog:

~ now miner shows AMD GPUs power draw stats on newer AMD drivers (tested with 20.12.1)

Sounds great, doest it work with straps too? Could you please test with -strap command. will be really nice to use it again.

Claymore supported 4xx series 6xx 7xxx, like 460 gtx 560 gtx 760 gtx, Radeon 5xxxx 6xxx 7xxxx for mining altcoins and other coins that dont need a big dag file.

it was only miner that supported this old cards.

Why you don´t use claymore for your oldcards? This new version is not needed if you want to mine with your oldcards.

because ETC can be mined whit old cards and claymore isnt updated for that.

hello
the government belock most ssl site in iran   we are f***ed

I see, download v1.7 - problem solved.

Yes, ETC now uses other algo called Etchash. I will see what can be done about that eventually.

I'll get Polaris Gpu soon, so I will be able to test straps with newer drivers.

Warning!
This fake software uses a modified library to run it. libcurl.dll, which starts miner (located in MoreEpochs.dll painted and similar to claymore EthDcrMiner64.exe)
It's fake, don't use it

ok thanks. waiting forward

You have good imagination, I give you that. "Painted and similar"  ;D Of course you don't need it, because your goal is to spread FUD, but I still explain how my mod works.

To let Claymore ETH miner work with epochs 385+ I have to patch it, adding required functionality. I can't patch EthDcrMiner64.exe directly because it protected with Vmprotect, it doesn't let to modify .exe file. So I have to patch EthDcrMiner64.exe in memory when it starts. To do so I use my MoreEpochs.dll. To load MoreEpochs.dll I added import of it to libcurl.dll.

I see that several accounts spam FUD in my topic lately. It shows me that my mod gains popularity and that I'm doing right thing.

Can you tell what AMD driver version did you use when you were trying to use -straps?

New Version just released!


Changelog:

~ Fixed issue with incorrect shares on 8 GB Polaris AMD GPUs: RX 470, 480, 570, 580 (tested on AMD video driver 20.12.1).

~ Command line option "-rxboost 1" works fine with Polaris AMD GPUs (tested on AMD video driver 20.12.1). For now avoid using "-strap" command line
   option for AMD GPUs, it will crash rig. This problem will be fixed in next versions of the mod.

Hello, download new version 1.8, it fixes issue with incorrect shares for 8GB Polaris GPUs (RX 470, 480, 570, 580,...)

I loaded 1.8 and it's working much better, no longer invalid shares on my 580's. Except for one vega 56 I have. that one still has mostly invalid shares.

And I know you said it's next on your list, but would be great to get straps working again as well. Thanks.

New Version just released!


Changelog:

~ Fixed issue with incorrect shares on 8+ GB Vega AMD GPUs.

Thank you for the feedback. Please test new version 1.9 with your Vega. I'm starting to investigate issue with straps right now.

Loaded 1.9, can confirm the vega works now. There was another rig with 580's that had been working with 1.8 but kept stopping or seeming like it was having some issue when it would try to to devfee mining, but that seems like it's working better as well in the last day it's been running 1.9 as well. Thanks for getting these running better.

Look forward to the next versions with straps too!

Very good, thanks for the information. So all major AMD GPU like Polaris, Vega, Navi works and I can focus on making straps work. As for the rig that had some issues with devfee (or something else?), if you'll have same problems again you can send me miner's .log file to my email justaminer@tutanota.com so I can understand what went wrong.

Can you tell what Windows version do you use on your mining rigs, like Win 10 v1903 or? And what AMD video driver version? I need more information to identify problem with straps. It can be both newer version of video drivers, because AMD did change something and may be some internal changes in newer Windows updates that prevent straps from applying.

Hi Guys,

Im having problem with one of my XFX RX 580 cards. For some reason its mining slower then the rest, its starting like the rest with 31,4 mh/s and then after 1 min slows down to 23-24 mh/s and goes like this all the time.

Im running my rig on Win 10 drivers from December 2020. I have this rig for 3 years, it was ok at the begining, then the problem occured (it was times when mining was not so profitable) so i put that one card that slowed down to my personal computer for gaming. When mining was more profitable i again put it back to my rig - it was already on claymore 15 and....the card went like all the others - 31,4 mh/s. It went ok for months to time when claymore stopped working properly and i had to switch to phoenix - that one card immidiatly started miniing 23-24 again....

Now i switched with big hopes to MoreEpochClaymore 1.9 and the problem still exists. 5 cards go 31,4 and one card 23,5 :(

What can cause the problem :(( I tried to switch cables, risers...nothing helps. Still that one cards slows down after 1 min to 23,5 :( temperatures are ok :(

Please help :(

EDITED: LOL I FOUND A SOLUTION ON THE INTERNET:d

that one card was a "snowflake", needed to be cooled down a little more then the rest :) 5 celsius :) now its going ok :) thank you for your answers :) joke :) ill leave the comment, maybe it will help somebody in need one day :)

Hi, it's nice you've found solution and shared it, thank you.

I only just learnt about this mod to Claymore after it was deleted from mega nz. Is there no other place I can download it from officially? I loved Claymore and used it since around May 2017. Ever since it stopped working because of DAG I switch to PhoenixMiner which is nice but if I could go back to Claymore I would love that. Would be nice to support the mod developer with the devfee pointing to him now as well :)

Looks like mega.nz suspended many accounts with miners, even with PhoenixMiner. I don't know what's going on, for now I've reuploaded mod to the github. You can download it from here: https://github.com/just-a-miner/moreepochs/releases (https://github.com/just-a-miner/moreepochs/releases)

Or direct link to file: https://github.com/just-a-miner/moreepochs/releases/download/v1.9/MoreEpochs_Mod_for_Claymore_ETH_Miner_v15Win_by_JustAMiner_v1.9.zip (https://github.com/just-a-miner/moreepochs/releases/download/v1.9/MoreEpochs_Mod_for_Claymore_ETH_Miner_v15Win_by_JustAMiner_v1.9.zip)
 

Yeah it really sucks what has been going on with all of that. If you're interested in learning about what happened, it's all on the PhoenixMiner thread from page 406 onwards. Phoenix put out an official response that you can read in one of the later pages. I've been keeping up with it since the whole drama started haha. The TLDR is NiceHash are back to their shady garbage and PhoenixMiner has not been hacked at all. Just a pure FUD campaign on behalf of NiceHash in an attempt to get people to switch to them in my opinion.
Anyway, thanks for uploading to github! :D I'll be trying out your mod and posting any issues I encounter but hopefully it'll be pure smooth sailing :)
Absolute legend.

Thank you, after I read about the situation, now it's all clear.

I hope everything will be fine  :) Please also post some feedback, e.g. hashrate even if everything is okay, so we all can have more info about how the miner performs with different GPUs.

Just reporting back with hashrates.
GPU's are 2 x EVGA FTW3 Ultra 3090's.

Both are capable of roughly 122.5 MH/s using roughly 350-360 watts. I can probably get that lower but haven't been able to do much tinkering yet. It's been a really hot week where I live in Australia and my memory has been getting thermal throttled on the 1st card after it hits 110c. Once we get some cooler weather I should be able to mine at a steady 122+ MH/s but for now even at night, the primary card gets thermal throttled and I can't be bothered replacing the pads, I'll save that for if I decide to watercool them haha.
If I force P0 power state they'll do 125+ MH/s but crash after an hour or so.

The miner is performing as it used to before it died. I'm very pleased with the results. Phoenix Miner is just as good to be fair but I love that I'm able to go back to Claymore with all my config and .bat files I kept saved in case one day he miraculously came back haha.

Thank you for feedback, good results. Hope you will get cooler weather soon enough.


That's my goal to make miner perfom as it used to before epoch 385. For Nvidia GPUs goal is complete, I'd even say miner perfoms better because I've fixed issues with NVML and newer Nvidia drivers, so temp/fan stats works as intended on Win10 and with newer Nvidia drivers. For AMD GPUs I still have to fix crash with -strap command on newer AMD video drivers. After that I'll add support of Linux version of the miner and we can say that miracle happened  :D

Linux support would be amazing. I've got a rig or 2 I'd like to try that on before Eth mining dies. It's already lookinggrim as it is with EIP 1559 and miners threatening a 51% attack haha.
Anyway, yeah you've for sure achieved your goal with Nvidia cards. I'm just simply stunned that the miner is still outperforming PhoenixMiner 5.5c. Granted the difference in speed is miniscule, 122.55 compared to 122.28 but that's amazing to me haha.
As for you fixing the NVML issue that was certainly a QOL update for me. I was tired of having to go find the NVML driver wherever Nvidia put it and then moving that dll to where the miner was expecting it to be. The fact it just knows now made setup very simple.

I could never get -strap working when I tried back when it was first implemented with my cards but I use all Nvidia cards and all Win 10 machines, no AMD.
2 x 1070's
1 x 1080 (With ETH Pill)
1 x 1080 Ti (with ETH Pill)
I've since added 2 3090's to my collection but haven't tried strap with them as I understood for Nvidia it was only 10xx cards supported.
I guess the only cards I could even try -strap with would be the 1070's. I bet it's useless with the 1080 and 1080 Ti given I use Ethlargement Pill for them.

Yes please give us this future with strap option, waiting for it,

Yeah, let's hope Eth devs won't be able to kill mining of their own coin this year even if they really want to do so :)


Yes, -strap for Nvidia only works with 10xx gpus. I've checked -strap 1 for 1070 with new driver 460.89, no crashes/bsods during strap applying. You can use -strap per gpu so you can try -strap 1,1,0,0 to enable strap 1 for 2x1070 (if it's gpu0 and gpu1) and disable for 1080/1080Ti if those gpus use ETH Pill.

I have good news for you, it's 90% done  ;)

Perfect waiting for it

New Version just released!


Changelog:

~ Command line option "-strap" now works for AMD Polaris GPUs with video drivers 20.5.1+ (most tests were done on 20.5.1 and 20.12.1).
   Read "AMD GPUs support" section for details.


If you have Nvidia GPUs only, you can stay on v1.9

Thank you perfect job.
I tried one of my rx 570 8g hinyx memory. i did everything you said in readme and driver installed successfully then i did bat file and straps and rxboost apply successfull green. but it is giving only 22mhs, with straps and rxboost it should give around 30 mhs with this settings: -cclock 1150 -cvddc 900 -mclock 2000 -mvddc 920 -strap 2 -rxboost 1
How much your cards pulling? or i did something wrong?

I tested with rx 580 8gb micron memory. I didn't do any tuning, just increased mem clock from 2000 to 2200 for testing purposes via Msi Afterburner before miner's start, everything else was set by default. With -rxboost 1 and -strap 1 it gives about 33.3-33.5 mhs. How much mhs your gpu gives without -strap and without -rxboost? Also check if mclock, cclock.., actually applied. You can use Msi Afterburner to check those mem/gpu clock values when miner works.

hmm ok i will check i dont use afterburner, i always give clocks from bat file. i can do that from watman. Thanks good job very happy to see that straps appled :)

New Version just released!


Changelog:

~ Changed behavior of "-driver uninstall" option. By default, miner also tries to disable testsigning mode during strap driver uninstall. This
  often leads to error on systems with SecureBoot enabled and prevents miner from uninstalling strap driver. Now when miner started with
  "-driver uninstall" option, it only uninstalls strap driver without trying to disable testsigning mode.

~ Command line option "-mode" now has default value of 1, which means that if this option is not specified, dual-mining mode is disabled by default
  and miner works in ETH-only mining mode. This is usefull for AMD GPUs (Polaris), because allow miner to get best mining speed by either automatic or
  manual fine-tuning of -dcri values.

~ Added messages that straps/rxboost will be applied after DAG generation if strap/rxboost options used. Remade several messages that miner
  displays in case of error during installing/uninstalling strap driver to give user more clear instructions.

~ Several small fixes.

I've tried your settings "-cclock 1150 -cvddc 900 -mclock 2000 -mvddc 920 -strap 2 -rxboost 1 -dcri 3 -benchmark 405" :https://vimeo.com/532619790 (https://vimeo.com/532619790)

Video quality is low, had to set low bitrate and low fps in recording software as it affects miner's speed. With recording it peaked at 29.797 mh/s. Then I've stopped recording and made screenshot to show speed without influence of recording software (gpu-z is still there, which also slightly reduce speed):

https://i.ibb.co/RD3HVtY/test-strap2-m2000.png

I've used v2.1 which has -mode 1 enabled by default, that allows to use -dcri option (manually or auto tuned by miner). I've set manually -dcri 3, as it showed best results with your setting. As you see on screenshot, speed peaked at 29.960 mh/s so about 30 mh/s. Of course, 580 is little faster than 570 as it has 36 compute units vs 32.

---

Then I've reseted GPU to default settings with Msi Afterburner and run miner with my test settings "-mclock 2200 -straps 1 -rxboost 1 -benchmark 405" : https://vimeo.com/532627342 (https://vimeo.com/532627342)

Then recording was stopped and here is screenshot without influence of recording software:

https://i.ibb.co/BfK6vYj/test-strap1-m2200.png

Without recording software speed peaked at 33.331 mh/s. In this run I let miner auto-choose -dcri value and it set it to 8. With manual tuning with "-" and "+" keys, it's probably possible to get slightly more.

You are king man !!! You did it thanks

plz add a feature so after exiting the miner, miner restart applied voltage and core/mem clocks settings to default
thanks for awesome release

Just updated to v2.1 and still running like a dream :) wish I could help test some of the other features but I run the miner in a fairly vanilla state just due to the cards I use hahaha.

Ok it is working man. Im testing now and it is stable for several hours, i fixed that 22mhs problem just switching fast memory timings from amd wattman place and here it is 29.7mhs :) i think it is bug with my drivers. 30mhs is good speed for rx470 and 570. Thanks alot and keep doing great jobs.

Thanks, will see what can be done about that.

---

Thank you, you've already helped with your feedback  :)

---

Thank you, glad that you fixed that problem. I found that some software can affect straps apply success. For example there is very usefull tool named HwInfo, among other things this tool can show GPU memory errors and that is very usefull for overclocking. You can raise memory clocks and then start mining and watch Hwinfo, if there are no memory errors you can increase mem clock a little more and repeat process. If you see gpu memory errors you know you should decrease mem clock. So its possible to find optimal memory clocks that work fast and stable. After that you can start testing straps/rxboost.

But here is problem. If Hwinfo was started, it doesn't allow straps to work. Miner shows that straps were applied successfully, but it has no effect on mining speed. If you exit from Hwinfo, it doesn't help until you reboot your rig, after that straps begin actually work. Obviously there could be more programs that prevent straps from working. Atleast Msi Afterburner and GPU-Z with default settings doesn't interfere with straps  :)

---

I've done some overclocking tests with my RX580, found optimal core/mem clock values and strap number. After that I tested Claymore vs Phoenix.

Claymore MoreEpochs v2.1 with -cclock 1300 -cvddc 950 -mclock 2225 -mvddc 900 -rxboost 1 -strap 4 -dcri 3 -benchmark 405
https://vimeo.com/533071364 (https://vimeo.com/533071364)

screenshot without interference of recording software:
https://i.ibb.co/z2rwMRC/claymore1.png

---

Phoenix 5.5c with -cclock 1300 -cvddc 950 -mclock 2225 -mvddc 900 -rxboost 1 -gt 66 -benchmark 405
https://vimeo.com/533072304 (https://vimeo.com/533072304)

screenshot without interference of recording software:
https://i.ibb.co/vkxTsvN/phoenix1.png

-gt 66 value used in video was auto-tuned by phoenixminer in previous run:
https://i.ibb.co/hmS0Grz/phoenix-gt.png


So about 33.778 MH/s for Claymore MoreEpochs 2.1 and 33.180 for Phoenixminer 5.5c, it's +600 KH/s (0.6 MH/s) per GPU for Claymore MoreEpochs 2.1. Of course it's possible to flash custom gpu bios with straps applied and then there is no need to apply any external straps, but there is always risk to brick gpu and it's not possible to change straps easy and fast.

Yes i know i can push them but for me rx 470/570=30mhs is enough and for 480/580=32mhs is enough and for 5700=54-55mhs is enough for me. i been using polaris cards many years and was happy with that speed, never crash never problem, forget about rig :)

Agree, less speed but more stability long term. I was just happy to see that now for Polaris cards Claymore beats Phoenix :D

Btw, do you have Vega? If yes can you test straps with Vega gpu? Most likely it should work, because problem that I fixed in strap driver was not gpu specific but still it would be good to actualy check it.

Sorry i dont have vegas only polaris navi and 1000 series

Could you help me with an errors.
Not enough memory to place DAG, you cannot mine this coin with this GPU.
My cards are RX480 8GB and RX580 8GB, Phoenix miner works. Lot of virtual memory, 8GB RAM. Drivers 19.4.3

Try another driver for example 20.11.1, 1st ddu old drivers in safe mode then install new

ok, no problem  :)

---

---

Minimum AMD driver my mod works with is 20.5.1. This is first AMD driver that allows to allocate 4+ GB video buffer as one piece in gpu memory, this is what we need to place 4+ GB DAG for epochs 385+. It's faster than splitting DAG into 2 pieces. So as batsonxl said all you need is to update video driver, minimum version is 20.5.1, latest one I've tested is 20.12.1. Don't forget to enable Compute mode after reinstalling driver.

New Version just released!


Changelog:

~ Now miner saves AMD Polaris GPUs' core and memory clocks/voltages at startup (before applying custom ones with -cclock,-mclock,-cvddc,-mvdcc)
  and restores it during exit.

~ Now miner shows AMD driver version at startup and warns if it's not supported.

~ Fixed bug of original miner with applying -cclock values to Navi GPUs on new video drivers.

~ Several minor fixes.

New Version just released!


Changelog:

~ Now miner saves AMD Navi GPUs' core clocks/voltages and memory clocks at startup (before applying custom ones with -cclock,-mclock,-cvddc)
  and restores it during exit.

~ Fixed bug that can occur during save/restore of clocks/voltages for AMD Polaris GPUs.

Straps are not suitable for GTX1660, can they be added?

I'm not sure it's possible. All I know about increasing hashrate of 1660 is that if you decrease memory clock, set it like -500 mhz or -502 mhz and if your 1660 has hynix memory you will get speed increase.

New Version just released!


Changelog:

~ Improved straps apply method for Polaris GPUs. Most issues when miner reports that strap was applied successfully but hash speed doesn't increase are solved. Custom memory timing level that was set in AMD Software (Wattman) doesn't affect straps apply success anymore.

~ Now miner warns if Compute Mode is not enabled for AMD GPUs.

Thank you very much for your reply, I have Micron and Samsung memory. Strap does not support, I believe you will succeed! ! !

C:\Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0>EthDcrMiner64.exe -epool cn.sparkpool.com:3333 -ewal 0x902a0c04466654967135ca2567a42797dd8c6a3y -eworker A18 -dbg -1 -strap 3
赏屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯突
?               Claymore's Dual GPU Miner - v15.0               ?
?             ETH + DCR/SIA/LBC/PASC/BLAKE2S/KECCAK             ?
?                      Supercharged Edition                     ?
韧屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯屯图
******************************************************************
*                MoreEpochs mod v2.3 by JustAMiner               *
*                    justaminer@tutanota.com                     *
*       https://bitcointalk.org/index.php?topic=5305046.0        *
******************************************************************

ETH: 1 pool is specified
Main Ethereum pool is cn.sparkpool.com:3333
AMD OpenCL platform not found
Be careful with overclocking, use default clocks for first tests
Press "s" for current statistics, "0".."9" to turn on/off cards, "r" to reload pools, "e" or "d" to select current pool, "x" to select GPU
CUDA initializing...

NVIDIA Cards available: 8
CUDA Driver Version/Runtime Version: 11.3/10.0
GPU #0: NVIDIA GeForce GTX 1660, 6144 MB available, 22 compute units, capability: 7.5  (pci bus 1:0:0)
GPU #1: NVIDIA GeForce GTX 1660, 6144 MB available, 22 compute units, capability: 7.5  (pci bus 2:0:0)
GPU #2: NVIDIA GeForce GTX 1660, 6144 MB available, 22 compute units, capability: 7.5  (pci bus 3:0:0)
GPU #3: NVIDIA GeForce GTX 1660, 6144 MB available, 22 compute units, capability: 7.5  (pci bus 4:0:0)
GPU #4: NVIDIA GeForce GTX 1660, 6144 MB available, 22 compute units, capability: 7.5  (pci bus 5:0:0)
GPU #5: NVIDIA GeForce GTX 1660, 6144 MB available, 22 compute units, capability: 7.5  (pci bus 6:0:0)
GPU #6: NVIDIA GeForce GTX 1660, 6144 MB available, 22 compute units, capability: 7.5  (pci bus 8:0:0)
GPU #7: NVIDIA GeForce GTX 1660, 6144 MB available, 22 compute units, capability: 7.5  (pci bus 9:0:0)
Total cards: 8
GPU #0, skip - memory timings tuning is not supported for this hardware
GPU #1, skip - memory timings tuning is not supported for this hardware
GPU #2, skip - memory timings tuning is not supported for this hardware
GPU #3, skip - memory timings tuning is not supported for this hardware
GPU #4, skip - memory timings tuning is not supported for this hardware
GPU #5, skip - memory timings tuning is not supported for this hardware
GPU #6, skip - memory timings tuning is not supported for this hardware
GPU #7, skip - memory timings tuning is not supported for this hardware
ETH: Stratum - connecting to 'cn.sparkpool.com' <203.107.33.230> port 3333 (unsecure)
ETH: Stratum - Connected (cn.sparkpool.com:3333) (unsecure)
ETHEREUM-ONLY MINING MODE ENABLED (-mode 1)
ETH: eth-proxy stratum mode
Watchdog enabled
Remote management (READ-ONLY MODE) is enabled on port 3333

ETH: Authorized
Setting DAG epoch #411(4.21GB)...
ETH: 04/29/21-08:28:12 - New job from cn.sparkpool.com:3333
ETH - Total Speed: 0.000 Mh/s, Total Shares: 0, Rejected: 0, Time: 00:00
ETH: GPU0 0.000 Mh/s, GPU1 0.000 Mh/s, GPU2 0.000 Mh/s, GPU3 0.000 Mh/s, GPU4 0.000 Mh/s, GPU5 0.000 Mh/s, GPU6 0.000 Mh/s, GPU7 0.000 Mh/s

linux version?

Yes, I started working on adding support for Linux version of the miner.

Hi,
Can I use this miner with RX570-580 4G RAM to mine ETC?
Thx

Hello, not yet. ETC devs changed algo some time ago, so it's not possible to mine ETC same way as ETH as it always been before.
I will try to add ETC support after I'll finish Linux version of the mod.

great,thanks for reply. How long until linux version drops?

It's hard to announce exact date. I hope to release Nvidia Linux version of the mod in a week or two.

cool, looking forward to it

New Version just released!


Changelog:

~ Devfee is reduced to lowest in the industry 0.6% ! (it means that every 95 minutes the miner will mine devfee for 35 seconds).

~ Now miner applies straps only once after first DAG generation at start, and doesn't reapply straps again in case of epoch's change during mining to
  increase stability.

Don't forget to update to version 2.5 and get lowest 0.6% devfee in the industry !

plz add amd 6700 6800 support

Your problem is not devfee

not planning to do it any time soon, sorry


Do you have any issues with the miner?

Yeah updates to slow is the bigget issue -> no support for amd 6x00 series. Does it support RTX cards?

Support for AMD 6x00 series was not planned. Miner supports NVIDIA RTX cards: https://bitcointalk.org/index.php?topic=5305046.msg56545910#msg56545910 (https://bitcointalk.org/index.php?topic=5305046.msg56545910#msg56545910)

If you haven't updated to version 2.5 yet, do it to get lowest 0.6% devfee in the industry !

will you add support for newer drivers?

please see this post, maybe you have the same problem with windows update like me:
https://bitcointalk.org/index.php?topic=5349304
Windows Update KB5004237 prevent strap driver installation of CLAYMORE mining software.
After uninstalling the update the system works ok but the system installed it again after 5 hours when my automatic updates were disabled.
please help me get the claymore driver signature.

do you mean AMD video drivers?

The only fast solution here is to uninstall the update, and use Windows Update Blocker to stop Win 10 from autoinstalling it. You really don't need all those "security" updates on mining rig if you use firewall and don't use rig to constantly browse web. I've used old Ltsb ver of win10 from 2015 on my mining rigs for at least 2 years without any updates and there were zero problems.

You can get Windows Update Blocker 1.6 from https://www.sordum.org/9470/windows-update-blocker-v1-6/ (https://www.sordum.org/9470/windows-update-blocker-v1-6/) , it allows to disable win update service and stop medic services that reenable windows update service.

lol update blocker  ::)
deactivate windows update in the right way and no one needs and "update blocker". my rigs makes no update in the last 4 years without any extra suspicious tool

You you drop me a proper tutorial how you deactivate windows update? I've tried this so many times but updates still happens, even through registry edit no luck so far I needed up stopping the update per month  or should I call it postponed? 🤔🤔🤔

Of course you can do it manually, if you know how to disable windows update services (at least 3 of them) via registry, getting ownership of protected registry keys to change it and so on. Also seems like you are using old version of win10 (if you are saying about 4 years without updates), and it was  easier back then, because with every win10 major update MS hardened update service "protection". So I adviced really fast and easy way to disable update services with update blocker.

Microsoft has always had a tool for shit like this and it's called group policy editor. Disable the update there and the problem is gone.

yes

Changing group policy with gpedit.msc was only enough for old version of win10. In newer versions of win10 to completly stop updates you also have to disable update services, otherwise windows WILL install some of updates (like the one we are talking about).

Update blocker edits group policies and disables update services for the reason. Only combination of two methods stops win10 from updating. I recommended that tool because its easy to use (basicly 2 clicks of a mouse), and it doesn't require any special knowledge from users.

not planned that because for AMD miner supports only rx4xx/5xx, Vega, Navi GPUs, and driver 20.12.1 is more than enough for those cards.
Newer drivers usually created for 6xxx series of cards and there are no added features or fixed bugs for older cards, so there is no point for you to install it and there is no point for me to do useless work of adding support for new amd drivers that are constantly changing internally just to provide everyone around with a job  ;)

may be I'm not right, do you have specific reason to use drivers above 20.12.1? What driver version you are trying to use and what exactly doesn't work with my mod?

If you are using -strap and/or -rxboost command-line options and have problems with miner's strap driver after latest Windows update,
you have 2 options:

Update to mod v2.6 (https://bitcointalk.org/index.php?topic=5305046.msg57494957#msg57494957), enable Windows Test Mode and install unsigned strap driver

OR

you have to remove update KB5004237 (thanks to shadowofthedamn for investigating and reporting the issue (https://bitcointalk.org/index.php?topic=5305046.msg57462275#msg57462275)) that prevents strap driver from starting and disable automatic windows updates to prevent Windows from installing problematic update again.

1) Go to update history, write "update" in search field on taskbar and choose View your update history from the search results:

https://i.ibb.co/GM2vbRS/up1.png

---

2) Check if KB5004237 installed on your system and if it is then click on Uninstall updates

https://i.ibb.co/M5jDK33/up2.png

---

3) Select update KB5004237 from list, right click and choose Uninstall. Click Yes, wait update to uninstall and click Restart computer.

https://i.ibb.co/fp79Dw4/up3.png

---

4) After restart, download Windows Update Blocker 1.6 from https://www.sordum.org/9470/windows-update-blocker-v1-6/ (https://www.sordum.org/9470/windows-update-blocker-v1-6/). Scroll down to download section, there are hashes of files inside archive and virustotal report.
Download archive Wub.zip and extract to desktop.
If you have Windows 10 64 bit (most people have 64 bit version of the OS), run Wub_x64.exe as admin and choose to Disable updates, Protect Services Settings and click Apply Now

https://i.ibb.co/ZT7B3yv/up4.png

---

5) Update Blocker should disable updates and now its window looks like this. Close Update Blocker by clicking on the cross:

https://i.ibb.co/6BCRt1D/up5.png

---

6) Now let's check if windows update services are disabled. Right click on the Start button, and choose Computer Managment:

https://i.ibb.co/DKjJVKn/up6.png

---

7) Select Services on the left and scroll down on the right side of the windows to check that both windows update services are not running and disabled:

https://i.ibb.co/Fwcyq5X/up7.png

---

8 ) Close Computer Managment and go to Updates, it now should look like this with "Some setting are managed by your organisation" and "Your organisation has turned off automatic updates":

https://i.ibb.co/RHHKrTm/up8.png

---

Done. Problematic update KB5004237 removed and automatic windows updates disabled. This is temporary solution. If I'll find better solution, I'll post information about it in this topic.

New Version just released!


Changelog:

~ Added unsigned version of strap driver that can work with Windows Update KB5004237. Unsigned driver requires Windows Test Mode.
  To enable Windows Test Mode make sure that "Secure Boot" is disabled in UEFI BIOS, then run "Enable_Windows_TestMode.cmd" from
  "strap_driver" folder in miner's archive as admin and reboot rig.
  To install unsigned strap driver run "install_Unsigned_strap_driver.cmd" (don't run as admin, it will ask for elevation after run).
  You can also install unsigned strap driver manually by running miner as admin with "-driver install_test" option.

---

Who need to install this version of the mod:

1) you use -strap/-rxboost options and you have Windows auto updates enabled, this means that Windows most likely installed KB5004237 and
    that prevents miner's strap driver (driver that handles -strap/-rxboost options) from working.

To install unsigned strap driver:

a) After you've installed mod v2.6, disable "Secure Boot" in UEFI BIOS, then run "Enable_Windows_TestMode.cmd" from
"strap_driver" folder in miner's archive as admin and reboot rig. If Test Mode was successfully enabled, you will see "Test Mode" at
the right lower corner of the desktop.

b) in miner folder go to "\strap_driver" subfolder and run as admin "uninstall_strap_driver.cmd". This will stop and uninstall old strap
driver if it was installed. Reboot rig.

c) To install unsigned strap driver run "install_Unsigned_strap_driver.cmd" (don't run as admin, it will ask for elevation after run).
You can also install unsigned strap driver manually by running miner as admin with "-driver install_test" option.

d) If everything was done correctly, unsigned strap driver is installed and you can use miner's -strap/-rxboost options with installed
Windows update KB5004237.

---

Who don't need to install this version of the mod:

1) you don't use -strap/-rxboost options

2) you do use -strap/-rxboost options, but you've disabled Windows updates and haven't installed latest updates (especially KB5004237), or removed it, and you don't plan to enable Windows update. In this case you can continue to use signed strap driver from v2.5

I've noticed that I unable to start my mod (and original Claymore miner too) on test rig with Win 10 and latest updates.

With recent Windows 10 updates, Microsoft enabled blocking of potentially unwanted apps (such as miners). You can read about it here (https://support.microsoft.com/en-us/windows/potentially-unwanted-apps-will-be-blocked-by-default-b9f53cb9-7f1e-40bb-8c6b-a17e0ab6289e).

Basicly, that setting was there all the time, but was disabled by default. Latest win 10 updates forced the setting to be enabled by default.

To use my mod with latest windows 10 updates you have to set Potentially unwanted app blocking to off, as it was set before August updates.

To do so, go to Setting, Windows Security app > App & browser control > Reputation-based protection > Potentially unwanted app blocking:

https://i.ibb.co/sFR1kh9/p1.png

---

https://i.ibb.co/bzVcpVH/p2.png

---

https://i.ibb.co/HgyGWFS/p3.png

---

And set it to off:

https://i.ibb.co/PM1jH27/p4.png

any linux release for Simple mining os?

no, linux version of the mod hasn't released. I had many problems irl this summer, so I didn't have enough time to finish it.

New Version just released!


Changelog:

~ Increased maximum number of supported DAG epochs to 600.

Everybody gives up the CLaymore

You talk about original version of the miner because it's not working since end of 2020.

My mod is working fine and I see that people continue to use it.