Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux)

[shYter90]

I am reading this thread every day
Some people said about different ways to change voltage on amd rx5xx series with new drivers, but I didn't see a lot of positive reviews.
So can someone please tell me the best solution to downvolt rx5xx series with new drivers (fix dag files)?
Thanks beforehande

[svdinu]

I am reading this thread every day
Some people said about different ways to change voltage on amd rx5xx series with new drivers, but I didn't see a lot of positive reviews.
So can someone please tell me the best solution to downvolt rx5xx series with new drivers (fix dag files)?
Thanks beforehande

This worked for me (4 x RX 580 Sapphire Nitro + LE):
https://bitcointalk.org/index.php?topic=1433925.msg21022569#msg21022569 with AMD Blockchain Drivers and OverdriveNTool.
Voltages 857 mV, P7 = 1172 MHz, P1 = 2250 MHz, 31.3 Mh/s

[NavyCuda]

I've been trying to get dual mining to work again.  I'm impressed by the performance with hbcc memory segmentation enabled.  I'm having an issue though.  If I do anything in windows outside of the miner, like clicking on the start menu and then clicking on the desktop again, locks up and I have to hard reset.

Code:

setx GPU_FORCE_64BIT_PTR 0
setx GPU_MAX_HEAP_SIZE 100
setx GPU_USE_SYNC_OBJECTS 1
setx GPU_MAX_ALLOC_PERCENT 100
setx GPU_SINGLE_ALLOC_PERCENT 100
EthDcrMiner64.exe -epool us2.ethermine.org:4444 -ewal 0xd909bB6EbAbBE96F46B07836267B7700C48894d2.C_VEGA_FE -epsw x -dpool dcr.suprnova.cc:3252 -dwal NavyCuda.vega_fe -dpsw x -allpools 1 -cclock 1137,1137 -mclock 1050,1100 -powlim 50,50 -tt 55,55 -fanmax 97,97 -fanmin 50,50 -cvddc 1000,1000

On single mining I get 40.2Mh/s on the 1050MHz HBM2 card. It is also my display adapter. On the previous drivers I could run -cvddc at 950,950 and stay stable while mining or playing gta v, with aug.23 blockchain I need -cvddc 1000,1000.  It's worth noting that similar mining results are possible with -cclock 852,852 but still runs the same temp and ~.5 Mh/s performance loss.

Dual mining I get 39.5Mh/s eth and ~1100Mh/s DCR. I've gone as high as 1200mV on each card to try and prevent them from crashing, but anything over 1050 and the computer goes straight to a hardcrash.

I don't want to go any lower on memory clocks because they have such a huge effect on performance.  I'm not that experienced in these things, however I have noticed that -cvddc has an effect on if a memory overclock will hold or not.  Not high enough and the memory will only hold 500 or 800Mhz.  GPU_0 is set at 1050MHz HBM2 to prevent artifacts and keep stability while mining.

I've gone back to single mining again for the time being but with such a small loss in performance it's worth it to me to dual mine.  Well if you have any thoughts it would be greatly appreciated, sleepytime!

[NDTVN]

Hello everyone,

i got 2gh/s. Can someone sugget me using claymore with nanopool or ethermine?

I don't know why i got lower profit with ethermine.

1 Month Ξ12.7 ETH in ethermine with 2gh/s

thank you

[Claymore]

Hi Claymore

Have you managed to replicate the issue with 12 gpu systems
issue: on version 9.8 if you have more than 3 amd cards in a rig of 11 or 12 gpus ethdcrminer hangs on Algorith ASM, if you start the instance with 3 amd cards and the rest nvidia then it starts mining. then you have to start another instance of ethdcrminer with the remaining cards, .
This issue was also experienced by a youtube channel Cryptomined where he had to have multiple instances on 9.8 running in order to use all the cards.
When all cards are selected after the cmd window hangs nothing but a hard reset solves the hang.
Hope that helps, if you need any more info or troubleshhooting with this regards you can let me know.
thanks
 

I don't have appropriate hardware to test 12GPUs configuration. You can try some different miner to check if it is more stable, because the reason can be not in miner. Also try "-gser 1" or "-gser 2" option.

[Claymore]

I think there is a software problem and I will explain why I think that.
I am running multiple 13 GPU rigs all of them equipped with ASRock H110 Pro BTC+ , all of them have a mix of RX 580 8GB and Nvidia cards, running 7/8 RX and rest Nvidia.
The Nvidia is farming an different software with zcash , the RX all use Claymore Dual Mining.
What is the problem:
When I use -di with more then 3 AMD cards , the software refuse to start. Windows doesn't frozen PC is still running but the Claymore remains frozen at GPU #x: algorithm ASM
If I use -di 012 for example (maxim 3 AMD Cards) the software runes perfect, so basic for 7 GPU , I need to run 3 Claymore miner's with -di 012 , di 345 , di 6 (any other with more then 3 cards will frozen the software)
What combination I try:
- If I disable ASM , it get frozen at POOL/SOLO version, so is not an ASM problem.
- I tested multiple Claymore Dual mining software, 9.8 / 9.7 etc, with different AMD drivers, mixed old drivers, latest block chain driver , or the newest one available on website, same story for all
- Also I tested on all miners gser 1 and gser 2 option from the manual , is not working.
- If I enable 3 AMD cards and 5 Nvidia cards, is working fine, the problem is just for AMD cards over 3 enabled
- I increase my virtual memory on one rig from up to 60GB , it didn't help.
- I tested the software with dual mining enable and ETH mode only, and any other possible combination from Claymore manual.
Note: the zcash software is turned off during this tests , so no other program to interfere with Claymore miner.
I appreciate your work and I hope you can find a solution for this problem, special that lot of motherboard vendors announces mining boards with lot's of PCIe slots, Ex ASUS 19 GPU slots.
Thanks,

Miner does not care about how many cards are used, it just calls OpenCL. So it can be drivers-related problem. What if you start Genoil miner? Yeah vendors announce mining boards with 12-19 PCIe slots, but drivers fail to support it properly, at least in Windows.

[dollos]

Hi Claymore

As you can see from the other user also the exact same issue.
i thought it might have been the board but he is using a asus board and i am using a biostar board.
exact identical problem.

i tried genoils but doesnt seem to use all gpus.
I tried very old drivers, i tried new drivers, i tried the crypto drivers, they all doing the same thing,
i am only using 4 AMD cards per rig of 12

if i only put 6 amd cards then its fine
if i put 6 amd cards with 6 nvidia then ethdcrminer gives the issue,
if i put 6 nvidia only then its fine also
if i put 4 amd and 6 nvidia then  also this issue.


all we are asking is just have a look at the code please.
Thanks

[don.ton]

Hi Claymore

As you can see from the other user also the exact same issue.
i thought it might have been the board but he is using a asus board and i am using a biostar board.
exact identical problem.

My Biostar is working right with 8 AMD + 4 Nvidia at the same time on Windows 10 x64.
Just two claymore runnning, first with the 8 AMD and the second with 4 Nvidia.

https://cdn.pbrd.co/images/GH7NHsF.png

[Claymore]

Hi Claymore

As you can see from the other user also the exact same issue.
i thought it might have been the board but he is using a asus board and i am using a biostar board.
exact identical problem.

i tried genoils but doesnt seem to use all gpus.
I tried very old drivers, i tried new drivers, i tried the crypto drivers, they all doing the same thing,
i am only using 4 AMD cards per rig of 12

if i only put 6 amd cards then its fine
if i put 6 amd cards with 6 nvidia then ethdcrminer gives the issue,
if i put 6 nvidia only then its fine also
if i put 4 amd and 6 nvidia then  also this issue.


all we are asking is just have a look at the code please.
Thanks

My miner is closed source so you cannot see its code, but you always can have a look at the code of Genoil's miner and fix it somehow if you have appropriate skills and think that the reason is in miner software. As soon as you make some other miner work properly, you can state that the reason of the problem is my miner.

[stanis_r]

Dear Claymore !
after updated with amd official blockchain driver i can't set clock and vddc with miner.  i had best settings for powerdraw and hashrate for my RX 580 Gigabyte gamining gpus with your miner with stable clocks and voltages before amd update. Now we have to use AB and then my power increase very high and clocks are unstable. when we will have update miner with fixed issued above ? Thanks in advance.

[vapourminer]

What's everyone's thoughts on mixing brands of cards?  2 x MSI 1060 + 2 x Asus 1060 for example.  Is this more work with managing OCs?

no problem.

in afterburner you can group similar cards and set them all at once, or not and set individually.

[NavyCuda]

Sorry to be pushy but does anyone have any ideas on my issue as posted on the previous page?

[nadrojcote]

I've been trying to get dual mining to work again.  I'm impressed by the performance with hbcc memory segmentation enabled.  I'm having an issue though.  If I do anything in windows outside of the miner, like clicking on the start menu and then clicking on the desktop again, locks up and I have to hard reset.

Have you tried disabling aero effects in windows?

[Claymore]

There are too many miners that use the software, and most of them use Linux .. so problems like this are isolated and will probably not gone be fixed, since they are not even seen as a problem by author. As simple as that we are invited to use a different software if we don't like what we get , or to resolve the issue our self a simple fuck off... even if some of us are new register doesn't mean we where not here from the begin.
If is not a software issue then why Claymore zCash works fine with all AMD cards, or Claymore dual mining works to launch simulate multiple scripts?
...btw similar issue describe by youtuber Cryptomined in one of his video.
Thanks,

1. Mining process involves several things: hardware, OS, drivers, miner software, internet connection, pool. To make sure that the problem is in miner, it is a good idea to confirm it by checking that other miners work as expected.
Same for other things, for example, to make sure that the problem is in pool, you should try to use some different pool to confirm it. That's why I tell you to try different miner, not because I'm going to ignore issues related to my miner.
For example, yesterday some people had probems with DAG #128, some of them were sure that the problem is in my miner, I said that it's a good idea to try different miner or pool to confirm the reason, because this problem can be due to miner or due to pool. I said it not because I don't care.
2. This issue can be related to miner or to drivers, or to hardware, but if other miners work, it will prove that the problem is in miner. Other coins require different OpenCL parameters so miners for ZEC may work fine. Also OpenCL drivers may have some bugs/limitations that allow to use all cards from different miner instances but fail if all calls are from single instance.
3. Note that right now I cannot investigate the issue directly because I don't have mb with 12 pcie, it was not necessary for me until today and it will take some time to buy it. So right now I can only suggest to confirm the reason of the issue by checking if other miners work. I can also check log files if you PM me them, that's all I can do.

[blackhwk77]

Odd, I run 1 instance of claymore with 8 AMD 570/580 and 3 GTX 1070 with zero issues on my Asrock/win10 rig.  I did set the options be less aggressive with DAG creation though.

Hi Claymore

Have you managed to replicate the issue with 12 gpu systems
issue: on version 9.8 if you have more than 3 amd cards in a rig of 11 or 12 gpus ethdcrminer hangs on Algorith ASM, if you start the instance with 3 amd cards and the rest nvidia then it starts mining. then you have to start another instance of ethdcrminer with the remaining cards, .
This issue was also experienced by a youtube channel Cryptomined where he had to have multiple instances on 9.8 running in order to use all the cards.
When all cards are selected after the cmd window hangs nothing but a hard reset solves the hang.
Hope that helps, if you need any more info or troubleshhooting with this regards you can let me know.
thanks
 

[NavyCuda]

I've been trying to get dual mining to work again.  I'm impressed by the performance with hbcc memory segmentation enabled.  I'm having an issue though.  If I do anything in windows outside of the miner, like clicking on the start menu and then clicking on the desktop again, locks up and I have to hard reset.

Have you tried disabling aero effects in windows?

I just realized it's probably driver related as I had no problem dual mining on older drivers other than having a noticeable drop in ETH mining, which is why I went back to single mining before.

[notbatman]

How do you get 12 GPUs in a rig? What board has 12 slots?

[blackhwk77]

Biostar TB250-BTC PRO - 12 pcie slots
ASrock H110 Pro BTC+ - 13 pcie slots
and now the ASUS B250 Expert Mining 19 pcie slots

I will say it can be a pain in the ass to track down issues with that many GPUs.

How do you get 12 GPUs in a rig? What board has 12 slots?

[greatauror28]

There are too many miners that use the software, and most of them use Linux .. so problems like this are isolated and will probably not gone be fixed, since they are not even seen as a problem by author. As simple as that we are invited to use a different software if we don't like what we get , or to resolve the issue our self a simple fuck off... even if some of us are new register doesn't mean we where not here from the begin.
If is not a software issue then why Claymore zCash works fine with all AMD cards, or Claymore dual mining works to launch simulate multiple scripts?
...btw similar issue describe by youtuber Cryptomined in one of his video.
Thanks,

1. Mining process involves several things: hardware, OS, drivers, miner software, internet connection, pool. To make sure that the problem is in miner, it is a good idea to confirm it by checking that other miners work as expected.
Same for other things, for example, to make sure that the problem is in pool, you should try to use some different pool to confirm it. That's why I tell you to try different miner, not because I'm going to ignore issues related to my miner.
For example, yesterday some people had probems with DAG #128, some of them were sure that the problem is in my miner, I said that it's a good idea to try different miner or pool to confirm the reason, because this problem can be due to miner or due to pool. I said it not because I don't care.
2. This issue can be related to miner or to drivers, or to hardware, but if other miners work, it will prove that the problem is in miner. Other coins require different OpenCL parameters so miners for ZEC may work fine. Also OpenCL drivers may have some bugs/limitations that allow to use all cards from different miner instances but fail if all calls are from single instance.
3. Note that right now I cannot investigate the issue directly because I don't have mb with 12 pcie, it was not necessary for me until today and it will take some time to buy it. So right now I can only suggest to confirm the reason of the issue by checking if other miners work. I can also check log files if you PM me them, that's all I can do.

Since I saw you reply here not too long ago, i'm just going to try and ask anyway...

Are you in the process of creating v9.9 that helps us set our clocks and voltages again in Claymore in unison with using blockchain drivers?

If yes, I bet the whole mining community that supports your software would be more then happy.

Thank you for creating your miner.

[darktrop]

I have multiple rigs running both Asrock H110 pro BTIC (has 13 pcie slots) and Supox motherboard that has 12 pcie riser slots. I cannot run more than 8 cards (nvidia 1070) on these rigs. I have an I3 core processor LGA 1151 16gb memory, vitual memory set to 16GB-22GB. Win 10 recognizes all the 12 or 13 cards but the moment claymore starts the system crashes. 8 GPU runs great. hashing 240 mhz.second with 8 GPU. Any thoughts on the command line text to split the GPU's between two claymore's?