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

[Mobsy]

I have troubles with my 6x rx 580 rig:
While the settings are ok and I did successfully flash the bios, seems that I still have to apply voltages using wattman.

Any idea why I am unable to set them from my start.bat?


timeout /t 15
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 eu1.ethermine.org:4444 -ewal mywallet.Rig1 -epsw x -r 1

-cclock 1200,1200,1200,1200,1200,1200
-mclock 2200,2200,2200,2200,2200,2200
-cvddc 900,900,900,900,900,900
-mvddc 900,900,900,900,900,900

Any help is much appreciated, thx.

Are you using claymore 10.5? If yes, sometimes I need to open and close the claymore bat for 2-3 times to set correctly the configs. Open, wait the last setting dag file and close. Do it for more 2-3 times.

YES - as in: you are right! had to restart it couple of times to get the settings. using claymore 10.4 atm, no reason to go for 10.5, at least from patch notes...
THX MAN!

[1715310609]

1715310609

[1715310609]

1715310609

[1715310609]

1715310609

[1715310609]

1715310609

[1715310609]

1715310609

[1715310609]

1715310609

[twicx]

Since switching to the new adrenaline drivers, I've found mining just about anything desperately unstable. Temperatures are higher, things are crashing frequently etc. But then on older drivers, I have to install amd settings to change workload to compute, and amd settings sucks.

Just wondering if anyone else is having these problems, and/or if someone could recommend a pre-adrenalin driver that supports more than 10 GPUs and doesn't suck?

[oubba37]

hello i am new in the mining
can i minne with my pc win32
i can't get the clymore for win32
can you help

[toastfacegrillah]

@kkourmousis

Hi there,

I believe i'm having your exact power spike issues, even with good quality PSU's Corsair HX1000i. I was wondering if you have any developments on it?

So far the ways I have alleviated it is by using the -gser 3 option and -lidag 3 to create artificial delays between the GPU initiation and power spike which works about 50% of the time. In windows 10 it will not cause the spike until the miner restarts at some point, then the OCP kicks in and rig switches off. If i use SMOS it will cut out at the exact same point each time. (when OC is being pushed to cards around the "creating GPU DAG buffer" part of Claymore boot)

- 5 x RX 580 and Ryzen 1200 draws 800w @ 150Mh/s Eth 2700h/s SiaCoin all under 70 Degrees C
- Corsair HX1000i PSU 1000w with 1100w Peak

-cvddc 850 -mvddc 875 -cclock 1200 -mclock 2180 Custom Bios, all Samsung VRAM

[muzuca]

Since switching to the new adrenaline drivers, I've found mining just about anything desperately unstable. Temperatures are higher, things are crashing frequently etc. But then on older drivers, I have to install amd settings to change workload to compute, and amd settings sucks.

Just wondering if anyone else is having these problems, and/or if someone could recommend a pre-adrenalin driver that supports more than 10 GPUs and doesn't suck?

Did you try blockchain Aug 23 drivers?
I’m using it in my rigs, but I only have 8 gpus in each.

[twicx]

Since switching to the new adrenaline drivers, I've found mining just about anything desperately unstable. Temperatures are higher, things are crashing frequently etc. But then on older drivers, I have to install amd settings to change workload to compute, and amd settings sucks.

Just wondering if anyone else is having these problems, and/or if someone could recommend a pre-adrenalin driver that supports more than 10 GPUs and doesn't suck?

Did you try blockchain Aug 23 drivers?
I’m using it in my rigs, but I only have 8 gpus in each.

I had success with those before, but I think they only support 8 or 10, I need 12. I know there was one shortly after that worked for me before, I think it was the November one, but I'm hoping someone here might know which one. I remember there was a few that if I didn't install amd settings, it would assume it was doing graphics and my hashrate would suck on anything it cryptonight. Ideally, I need a stable pre-adrenalin one, where I don't need to install amd settings.

I just wanna go back to ethash on sgminer

Edit: now that I think about it, maybe those blockchain drivers were multi gpu. I'll give them a go!

[nzalog]

Nice job on 10.5, finally seems rock solid. 

[kkourmousis]

@kkourmousis

Hi there,

I believe i'm having your exact power spike issues, even with good quality PSU's Corsair HX1000i. I was wondering if you have any developments on it?

So far the ways I have alleviated it is by using the -gser 3 option and -lidag 3 to create artificial delays between the GPU initiation and power spike which works about 50% of the time. In windows 10 it will not cause the spike until the miner restarts at some point, then the OCP kicks in and rig switches off. If i use SMOS it will cut out at the exact same point each time. (when OC is being pushed to cards around the "creating GPU DAG buffer" part of Claymore boot)

- 5 x RX 580 and Ryzen 1200 draws 800w @ 150Mh/s Eth 2700h/s SiaCoin all under 70 Degrees C
- Corsair HX1000i PSU 1000w with 1100w Peak

-cvddc 850 -mvddc 875 -cclock 1200 -mclock 2180 Custom Bios, all Samsung VRAM

Hi,

I have been mining for ~2 months now with 0 problems. How? Well, when the first AMD drivers that had the "Compute" mode came around (I think it was 17.11.1. Now I am on 17.11.3 and I am thinking of installing the latest but you know how it goes: if it works, don't fix it), they made the difference.

BUT I did not install them without problems. After installation, my rig worked for ~1 day with no problems, but then the miner crashed and it would not start again! It got stuck at the DAG file generation.

After some thought, I took the long way and formatted my drive. Clean install on everything, latest drivers etc. And... the same thing! 1 day of 0 problems and then the miner crashed. After research, I found out that it was the damn windows update that messed something up (at that time, the creator's fall update had just been released -I think it is relevant, that's why I am mentioning it). So, I performed another format, made sure I blocked Windows Updates by any possible way I could and voila: 2 months of no headaches.

So, if you haven't tried it already (because you are bored like I was, or something like that) a clean Windows installation will probably solve your problems. Search around for ways to block Windows Updates (it is a bit tricky on Windows 10) and whenever a new driver comes out resist the urge to install it if your system works well.

A final tip: just after a clean Windows install, proceed to install all your drivers EXCEPT your GPU drivers. Also, install all your software (like TeamViewer, HWInfo etc), perform any system configuration you want but do not install any gpu drivers or any Claymore version. Once you have done that, create a restore point AND do a full backup (there is a Windows 7 backup utility within Windows 10) to an external drive. That way, you will not have to perform a clean install again if something goes wrong with your drivers. That saved me a couple of hours so far and will definitely save me a lot more in case something goes terribly wrong in the future.

I Hope I helped

[Viaggiatore]

Hello, I have a rig with 2 RX 580 8gb and 1 RX 570 4gb running very smooth with 10.2. Today I added a RX 470 4GB and I got this error on this GPU "opencl error -4 cannot create DAG file for GPU3". What could be the issue?
Thanks a lot

[madpapa]

Hello, I have a rig with 2 RX 580 8gb and 1 RX 570 4gb running very smooth with 10.2. Today I added a RX 470 4GB and I got this error on this GPU "opencl error -4 cannot create DAG file for GPU3". What could be the issue?
Thanks a lot

Increase virtual memory.

[Viaggiatore]

Hello, I have a rig with 2 RX 580 8gb and 1 RX 570 4gb running very smooth with 10.2. Today I added a RX 470 4GB and I got this error on this GPU "opencl error -4 cannot create DAG file for GPU3". What could be the issue?
Thanks a lot

Increase virtual memory.

I have 24gb as initial and maximum size, what do you suggest? Thanks

[Merutsu]

With last versions(10.4 and 10.5) my both rigs(rx570 and rx580 cards) are crashing to bsod with 0x000000ea code randomly. Minidump shows it caused by dxgkrnl.sys and atikmdag.sys.
Have no this issue on 10.2 version.
//using win10 + blockchain driver.

[strelly83]

Hi, since last week I've some problem with the latest versione of Claymore. I use dual mining ETH + Decred with Ethermine pool and DCR suprnova and I'm having a high number of "Stale Shares".
If I delete Claymore folder and "reinstall" (unzip) an restart the rig the problem resolves for about 12 hours...do you have any suggestions?

thank you!

[Merutsu]

With last versions(10.4 and 10.5) my both rigs(rx570 and rx580 cards) are crashing to bsod with 0x000000ea code randomly. Minidump shows it caused by dxgkrnl.sys and atikmdag.sys.
Have no this issue on 10.2 version.
//using win10 + blockchain driver.

Are they overclocked? I've seen an increase in errors aswell, used HWInfo64 to reduce memory errors and had no more errors since then.

mem clock 1975mhz on all cards, gpu memory errors 0 for all cards, have tested on 10.2, 10.4 and 10.5 versions. For now i switched back to 10.2, this one is stable.

[Viaggiatore]

Hello, I have a rig with 2 RX 580 8gb and 1 RX 570 4gb running very smooth with 10.2. Today I added a RX 470 4GB and I got this error on this GPU "opencl error -4 cannot create DAG file for GPU3". What could be the issue?
Thanks a lot

Increase virtual memory.

I have 24gb as initial and maximum size, what do you suggest? Thanks

Did you overclock the 470, try to run from stock first.
Did you reinstall the drivers? Try to reinstall them first.

Uhm, just saw that crossfire activated by itself-> voltages and clock settings where not taken by the 580s -> the 470 got the wrong values... I'll try with right voltages and clock values, thanks

[HarryTheHarry]

dear all,

currently im running ubuntu 16.04 lts rig, with 3x rx 580. the problem is ethash rate on dual mining (solo as well). on modded bios cards i hardly squeeze 21mh.
no matter what are the mclock/cclock values hashing varies 18-21mh/s. those values are reachble only wits asm 2 mode, solo eth mining does not help here. i really wasted lots o time no significant improve.
however at zec mining i got 330+hs per gpu which is oustanding, but what the heck can i do with eth?

[rusan]

dear all,

currently im running ubuntu 16.04 lts rig, with 3x rx 580. the problem is ethash rate on dual mining (solo as well). on modded bios cards i hardly squeeze 21mh.
no matter what are the mclock/cclock values hashing varies 18-21mh/s. those values are reachble only wits asm 2 mode, solo eth mining does not help here. i really wasted lots o time no significant improve.
however at zec mining i got 330+hs per gpu which is oustanding, but what the heck can i do with eth?

I had same problem on linux. And it was new core and need to reinstall drivers.

[Caiowulf]

Hi,
I am Mining Ethereum and Decred (or just Ethereum) with ethermine (and suprnova for Decred).
I have a GTX 1070 (only one, it is my first try on this) in my gaming computer. So I downloaded claymore and followed a few tutorials and explanations about configuring my miner.
I have a good 27.7 Hashrate with stock settings and after OC and some tweaks I have reached 30.5 Hashrate in solo mining.
I am using Windows 8.1 Pro.

My problem is that, for example with the stock settings, my 27.7 hashrate only stays if I am somehow using the computer at the same time that I am mining. If I let it mining alone, closing every other program, etc, my Hashrate drops, in 3 or 4 minutes, to 18. And this happens always, after I have been AFK.
After that, the miner remains in 17~19 for hours, being consistent and stable, but at 18 and not at my 27 peaks.
But, if I stays using the computer, the miner remains in 27~28 being consistent and stable.

Do anyone have any clue about what can be happening ?
I already tried tweaking power options and can't make it change a bit. It is still the same (may be it is still the power options and I lacked on changing the proper setting, i don't know).

Thanks a lot. I am pretty knew on mining.

[jrr]

v10.4:

- fixed issue with rejected shares in ETH+SIA mode for Vega cards.
- improved watchdog for temperature/fan management thread.

I'm have a problem when running it as a Task, the gpu infos(temperature and fan speed) don't show on web monitor. If I run normally by clicking in the start.bat, it works as expected.

It is not a problem in just the 10.4 version, I have used the 10.3 for a while and it never showed the gpu info when running as task.

I have the problem in two computers, the same config, both with one with XFX Rx 580 8GB or other with 4GB;

The configuration on the task or in the start.bat file is the same, I tried -wd 1, which is the default, but didn't work.

Someone else has the same problem?

The same problem with the last version 10.5

[HarryTheHarry]

dear all,

currently im running ubuntu 16.04 lts rig, with 3x rx 580. the problem is ethash rate on dual mining (solo as well). on modded bios cards i hardly squeeze 21mh.
no matter what are the mclock/cclock values hashing varies 18-21mh/s. those values are reachble only wits asm 2 mode, solo eth mining does not help here. i really wasted lots o time no significant improve.
however at zec mining i got 330+hs per gpu which is oustanding, but what the heck can i do with eth?

What RAM brand do you have and what memory straps did you flash?

I remember when flashing UBER 3.1 timing on my samsung memory my hashrate went down vs the 1750 timings

i got xfx with samsung, which i managed to flash 1:1500 ^ and 2:1500 ^
2x sappphiro with micron first 1:1650 ^ and second 1:1750 ^

I had same problem on linux. And it was new core and need to reinstall drivers.

what you mean by new core? as per drivers i have recomended amdgpu-pro-17.40-483984, which were reinstalled already couple of times