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

[LaskoFiasco]

I'm trying to resolve an issue with remote management in Windows.

I've tried remotely managing my miners using Remote Desktop, TeamViewer, and Claymore's Built-In Remote Management, but unless I physically connect a monitor to one of the GPUs, the miner software (EthDcrMiner64.exe) will only display GPU temps and make core/voltage changes to 1 GPU (the one with the HDMI Emulator Plug).

The odd thing is that I can access the Temperatures, Voltages, and Frequencies of all cards within WattMan using this same setup, but Claymore will only read the values from 1 card (all 7 cards do hash just fine). It would be far more convenient to manage GPU settings within config.txt itself as WattMan is tedious to configure Core & Memory clocks and Voltages for 7 cards on 4 rigs.

My setup is as follows:

• Windows 10 64-bit
• MSI Z170A Gaming M5
• 7x - Sapphire Nitro (non-Plus) Radeon RX 480 8GB
• AMD Catalyst 16.9.1
• Headless HDMI Emulator connected to the 1 GPU that is detected (without this I cannot access WattMan).

Has anyone else run in to a similar situation? Or alternatively, I'm open to switching to Linux, however I have not found a way to underclock memory voltages within Ubuntu.

[1714050708]

1714050708

[1714050708]

1714050708

[1714050708]

1714050708

[1714050708]

1714050708

[Toxic_Cat]

Miners with big farm already updated from 8.1 to 9.6? What is final results?

I have strange results. Hashrate is unstable. First GPU#0 (with HDMI Dummy) always drop from 27.6 to 24Mh/s
Using ASM #2 on RX 470/480/580 4G

[win33]

I have r9 280x mining on dwafpool current hashrate @ 12.5mh mining with claymore dual ethereum is that normal?
or some driver issue?
used to have around 24 using ethminer.

also i have gtx 970, hashrate @ 3.4mh , how to get it back to like 18mh like in ethminer



what are the factor to determine mining hashrate?
server/internet speed/ does it play a part?

What??? From where to 24 mhs on r9 280x cards?
I have only 12-14 mhs on ethminer.

[Darwin84]

I have a strange card and maybe someone can help or is familiar with a 0.00mhs hashing card.

I isolated a troubled 7970 and moved it to a separate computer. It runs windows ans linux fine, I just watched a 2hr movie and rendered some 3D scenes.  However when start up the 7970 it mines at 0.00mhs
and keeps on going forever at 0.00.  Has anyone had this happen before ?  It was hanging up in a rack of 5 and causing claymores to crash but alone it doesn't even produce a hash yet it runs and keeps on running getting the requests and works a regular video card. I have already tried several flashes all with the same result.

[yclycl]

ethminer.exe -U -S eu2.ethermine.org:4444 -SP 1 -O 0xa30ec91808321f158022956f742e62157f4ad0dd.glavni

you can try also --cl-local-work 128 --cl-global-work 8192

I have no %10 speedup with gtx1060, I have %2.5. I use gtx1060 9gbps edition, unfortunately all of them have performance problem for gtx1060 9gbps edition

try this!

@echo off
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
ethminer.exe --farm-recheck 2000 -U -S eu1.ethermine.org:4444 -FS us1.ethermine.org:4444 -O yourwallet.name


it work wonderful with mine 6x 1060 3gb giving 130mh(claymore) to 145/150mh (ethminer!!)


if you want to mine with amd just put -G to -U

Unfortunately, it doesn't work too. I could reach 19.77 mh maximum as stable. I suppose that algorithms must be updated for 9gbps edition, because another people that have 9gbps edition have this problem

[CrayjeksAjek]

MSI R9 390X
connected directly on pcie 16 slot, undeclocked, overvolted. Only below 1000 MHz (1080 stock) and over +35 mV i get any kind of stability (hour or two)

Keep getting GPU 1 hangs in OpenCL call error.
Windows says "Default Radeon WattMan settings have been restored due to unexpected system failure".

Current drivers 17.1.1, this hapened with 15.12 too.

Furmark QHD + 4x MSAA crashes when reaches around 40% of the test.


Is this card candidate for rma?
Anyone had anything similar happen?

Rest of the rig works fine no matter the driver version.

Need help.

I use Crimson for clocks and msi afterburner for fans. Fixed my opencl issue I was having for a week.

[ssnova]

Hey guys, I'm about to recalibrate my rig to try out this dual miner.  I read through some of the posts but not all of them, and I was wondering is Crimson 15.12 still the best driver for R9 Fury's (and R9 Nano.. seeing that the Radeon Pro Duo is 2x Nano's)  on Windows 10?


Also, is it safe to download legacy drivers from guru3d?

http://www.guru3d.com/files-details/amd-radeon-software-crimson-15-12-driver-download.html

I believe 16.3 is the best driver for R9 Fury.

Do you know from first hand experience with this miner and the specified cards? I only ask because Claymore states on the first page:

For AMD cards, Catalyst (Crimson) 15.12 is required for best performance and compatibility. You can get very bad results for different drivers version, or miner can fail on startup.

I use his miner for Zcash, and that is infact the case, 16.32 is the preferred driver for Claymore 12.5 Zcash, but this miner is a different beast, and I didn't see any mention of the R9 Fury in his OP, but I did see that he mentioned 15.12... in Zcash he mentioned the 16.xx drivers, which I use.

I have quiet a few r9 furys, over 120 that is if you have r9 fury and running claymore, doesn't it say use 16.3 for best results?

Ah nice, good to hear that your setup is getting decent results, 120m/s with how many Fury's? 

If you look on the first page though, there is no mention of 16.3 for best results, he mentions 15.12 for "older" cards, and latest for RX, I just rechecked.  On the Zcash miner, yes, he states 16.3 for best results with R9 Fury's, which is what I use right now for Zcash.

[Andrey123]

Hello!
In the morning I went into the miner, the 133rd era was downloaded and after the hasht became 0!
What does it mean?
After the reboot, the hashed returned.

[Eliavres]

@Claymore

I'm sticking with v9.5 for now. My hashrate dropped .5 MH/s on v9.6 with a MSI 1060 6gb (samsung).

[TechPark]

Hey guys, help to troubleshoot, on ethermine.org my ETH rate is well below reported though SIA is OK. Should i switch pool or look somwhere into my system? Win 10 latest version, latest drivers, on wifi:


WiFir esults:

Anyone can help?

Q:  Should i switch pool or look somwhere into my system?
A: Look somewhere in your system.

[TechPark]

ethminer.exe -U -S eu2.ethermine.org:4444 -SP 1 -O 0xa30ec91808321f158022956f742e62157f4ad0dd.glavni

you can try also --cl-local-work 128 --cl-global-work 8192

I have no %10 speedup with gtx1060, I have %2.5. I use gtx1060 9gbps edition, unfortunately all of them have performance problem for gtx1060 9gbps edition

try this!

@echo off
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
ethminer.exe --farm-recheck 2000 -U -S eu1.ethermine.org:4444 -FS us1.ethermine.org:4444 -O yourwallet.name


it work wonderful with mine 6x 1060 3gb giving 130mh(claymore) to 145/150mh (ethminer!!)


if you want to mine with amd just put -G to -U

130, 145 and 150 are reported hasrates. What are your effective hash-rates at the pool side? Is there 25-30% difference as well?

[Pair]

Hey,

I am having the issue that Claymore doesnt detect my GTX 970 GPUs.

I already tried to change to Driver 352.84 but still having this issue.

OS is Win 10

Nobody has an idea to fix that?

@Claymore maybe you can help me?

[dante85]

ethminer.exe -U -S eu2.ethermine.org:4444 -SP 1 -O 0xa30ec91808321f158022956f742e62157f4ad0dd.glavni

you can try also --cl-local-work 128 --cl-global-work 8192

I have no %10 speedup with gtx1060, I have %2.5. I use gtx1060 9gbps edition, unfortunately all of them have performance problem for gtx1060 9gbps edition

How to use this on suprnova pool? With the above command (replacing the pool address by eth.suprnova.cc:5000) I get errors: "Read response failed: End of file".

Seems like ethmine.org is not very interesting for me mining with one GPU => https://ethpool.freshdesk.com/support/solutions/articles/8000028994-when-will-i-receive-my-payment-on-ethpool-org-
It states =>  "the miner who contributed the most work to the pool will receive the full reward of the next found block"

Does that mean the Unpaid Balance on the overview page https://ethermine.org/miners/3Af998B30F4b4AC0d170dCe3CcBa4De9b8b9f494 is not going to be paid?

[SS2006]

So this just happened: https://github.com/ethereum-mining/ethminer/pull/18

Getting extra 2 mh/s on my GTX 1070, and reportedl the GTX 1060 sees even more improvement (see here: http://cryptomining-blog.com/8852-new-optimized-ethminer-for-nvidia-geforce-gtx-1060-gpus/)

Thing is I love mining Siacoin at the same time so I'm sticking with Claymore's for now, any chance these optimizations by David Li can be included in next version?

[Jacknife]

ethminer.exe -U -S eu2.ethermine.org:4444 -SP 1 -O 0xa30ec91808321f158022956f742e62157f4ad0dd.glavni

you can try also --cl-local-work 128 --cl-global-work 8192

I have no %10 speedup with gtx1060, I have %2.5. I use gtx1060 9gbps edition, unfortunately all of them have performance problem for gtx1060 9gbps edition

try this!

@echo off
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
ethminer.exe --farm-recheck 2000 -U -S eu1.ethermine.org:4444 -FS us1.ethermine.org:4444 -O yourwallet.name


it work wonderful with mine 6x 1060 3gb giving 130mh(claymore) to 145/150mh (ethminer!!)


if you want to mine with amd just put -G to -U

130, 145 and 150 are reported hasrates. What are your effective hash-rates at the pool side? Is there 25-30% difference as well?

I think the effect should be the same.

[TechPark]

ethminer.exe -U -S eu2.ethermine.org:4444 -SP 1 -O 0xa30ec91808321f158022956f742e62157f4ad0dd.glavni

you can try also --cl-local-work 128 --cl-global-work 8192

I have no %10 speedup with gtx1060, I have %2.5. I use gtx1060 9gbps edition, unfortunately all of them have performance problem for gtx1060 9gbps edition

How to use this on suprnova pool? With the above command (replacing the pool address by eth.suprnova.cc:5000) I get errors: "Read response failed: End of file".

Seems like ethmine.org is not very interesting for me mining with one GPU => https://ethpool.freshdesk.com/support/solutions/articles/8000028994-when-will-i-receive-my-payment-on-ethpool-org-
It states =>  "the miner who contributed the most work to the pool will receive the full reward of the next found block"

Does that mean the Unpaid Balance on the overview page https://ethermine.org/miners/3Af998B30F4b4AC0d170dCe3CcBa4De9b8b9f494 is not going to be paid?

the links your have pasted above are from 2 different pools. 2 different pools have 2 different payment structures. And none of the above related to performance issues with your card. Look for a solution locally.

[TechPark]

ethminer.exe -U -S eu2.ethermine.org:4444 -SP 1 -O 0xa30ec91808321f158022956f742e62157f4ad0dd.glavni

you can try also --cl-local-work 128 --cl-global-work 8192

I have no %10 speedup with gtx1060, I have %2.5. I use gtx1060 9gbps edition, unfortunately all of them have performance problem for gtx1060 9gbps edition

try this!

@echo off
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
ethminer.exe --farm-recheck 2000 -U -S eu1.ethermine.org:4444 -FS us1.ethermine.org:4444 -O yourwallet.name


it work wonderful with mine 6x 1060 3gb giving 130mh(claymore) to 145/150mh (ethminer!!)


if you want to mine with amd just put -G to -U

130, 145 and 150 are reported hasrates. What are your effective hash-rates at the pool side? Is there 25-30% difference as well?

I think the effect should be the same.

What makes you think that?

[BrianH]

I found some messages that slowdown for bigger DAGs is related to my miner only, and sgminer is ok. I would like to see some proofs instead of rumors from youtube.
So if someone can post the command line for sgminer for some future epoch (e.g. #150 or higher) here and the link to sgminer binary that should be used for tests, it would be good and we all could confirm it.

I confirmed the slowdown using Genoil miner, although the dropoff was not as severe.

For comparison, my cards currently (DAG Epoch ~130) run at around 83MH/s using Claymore's miner.

Results (3 x RX 480s):
DAG Epoch 0: 78.3MH/s
DAG Epoch 130: 77.6MH/s
DAG Epoch 199: 45.7MH/s

The command was something like this to test DAG Epoch 130:
ethminer.exe -G -M 3900000

"If using Genoil’s ethmienr with the -M option to benchmark you are not setting the DAG epoch number, but the number of block to test against… so make sure you set the right block number, each DAG epoch is 30K blocks, so just do the math."
http://cryptomining-blog.com/8822-ethereum-hashrate-drop-for-radeon-rx400rx500-gpus-is-incoming/#comment-5667

[sweetydady]

I left a comment on BBT video saying Claymore said don't believe those youtube guys, GSMiner has the same hashdrop problems with new epochs

hi replied

Bits Be Trippin'
 9 hours ago
SGMiner (Linux) using OpenCL 2.0 does not have the same issue after EPOCH 150+

[mjmeans]

This seems terribly poor for dual RX480 cards (VisionTek RX 480 8 GB OC + XFX RX 480 8 GB OC):

ETH - Total Speed: 29.125 Mh/s, Total Shares: 15, Rejected: 0, Time: 00:26
ETH: GPU0 16.488 Mh/s, GPU1 12.636 Mh/s
 SC - Total Speed: 873.735 Mh/s, Total Shares: 14, Rejected: 0
 SC: GPU0 494.650 Mh/s, GPU1 379.085 Mh/s
GPU0 t=86C fan=41%, GPU1 t=80C fan=36%

I'm using

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 eth-us-west1.nanopool.org:9999 -ewal myaddr.rig1/myemail -epsw x -dcoin sia -dpool sia-us-west1.nanopool.org:7777 -dwal myaddr/rig1/myemail -dpsw x -ftime 10

When I mine ETH only I get 48 MH/s. I have disabled crossfire in Radeon settings with the expectation that the different brands of cards would have issues with crossfire.

Am I doing something wrong? Should I be getting better than that?