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

[chronek]

I changed for new amd official win7-crimson-relive-beta-blockchain-workloads-aug11 driver and patched with atimdag-patcher 1.4.5 (bios modded only mem straps),
Claymore 9.8 crashing at start (just after listing cards and "POLO/SOLO version" caption, in log after "start building OpenCL program for GPU 0..").
It crash with stock settings or changed with OverdriveNTool 0.1.3....

Have anyone similar problem? Is that driver problem, claymore problem, other problem? was working before - 5x rx480, win7 x64

I had the same issue, I had to use DDU then reboot install the BETA Driver and then reboot, then pixel patcher then reboot again, after that it worked fine,

I did 2 times - run in safe mode, run DDU, reboot, install official amd beta driver, reboot, use pixel patcher, reboot, claymore 9.8 crash when try build opencl program...
I have windows 7, saw couple posts that there is problem on that os, is amd working with microsoft for push windows 10? (warning tinfoil theory)

[1714247813]

1714247813

[1714247813]

1714247813

[1714247813]

1714247813

[1714247813]

1714247813

[StephdeLyon]

hello
i have a big problem with the last version and all over miners. works with 3 cards and when i want to put 4 gpu cards on my rig, it crashes...
all the same cards, Under win10 fresh installed, AMD driver 17.4.4, stock cards, psu 1200w seasonic prime on asrock H81 rev2

risers are ok because:
- gpu 1+2+3 => ok
- gpu 1+2+4 => ok
- gpu 1+2+3+4 => ko...

claymore start well and detect the 4 gpu's but crashed before to start mining, just after GPU detect.

please help me !!

- Make sure the card is not sharing power with another card
- Make sure each riser not sharing its power with another riser
- Run all cards on stock BIOS and stock seetings
- Uninstall MSI AB or similar software
- Run DDU in safe mode
- Reinstall AMD driver, the latest
- Run Claymore with minimal settings -cvdcc 1000 -mvdcc 1500

Share results and screenshots from GPU-Z

hello,
i did everything and its still the same problem... im running now Under 4 cards and it's crashing.
Under GPUZ i don't see any problem and all the cards are detected. no time to post screenshots for the moment.
any idea ?

i put under 2 cards, i've got another problem about drivers, perhaps it's my problem, with 2 cards, on pciexpress 1 and 2,
one cards is Under Crimson 17.7.2 / whql /win10 64 drivers (opencl ticked)
and the second one is only whql / win10 64 drivers (opencl not ticked)

Did you run DDU? be honest...

of course i did it !
i had to go to my job this morning, not enought time sorry.
here is the screenshot with one card, it seem that there an exclamation behind my card. Conflict between windows drivers and AMD driver ??
do i have to use DDU each time i add a card and re install crimson AMD drivers 17.7.2 ?
https://www.noelshack.com/2017-33-3-1502867786-capture.png

i tried to use simplemining and it does not work well at all... only 10mhs with 3 cards and gpu core often low at 300 instead of 1236, really strange. I had the 300mhz issue yesterday and it was afterburner causing the issues. I pressed the reset button in afterburner and the clock went back up. Some say online to remove afterburner and reinstall it.

I believe you have two problems. First of all you need to reset all you afterburner settings if you are using it. Or if you are using any other overclock software just uninstall it until you get things sorted.

Then your 4th card is probably not showing up because of the motherboard bios. You need to make sure above 4g decoding is enabled. Some mobo brands call it something else. Some don't support it.

I use them stock and no msi afterburner installed.
I have a h81 asrock rev2, i will look for this,  but my 4th card is recognized but it crashes. Sometimes its difficult, i have to change pciexpress port to be recognized but it works.

[don.ton]

Is there anyone here that tried to undervolt and overclock using Claymore's miner (any version) with the new official Beta drivers?

I tried it. Doesn't works.  
For me the same with Wattman or AB, doesn't works for undervolting.

[MrJaekin]

Has anyone tried the new beta drivers with the VEGA 64 card?

[johndoe2]

Hello. Thank you to all the devs for creating such a great miner.
I was trying to dual mine with siacoin and it kept giving me this error.  Server: bind failed with error: 10048
I added the last two lines from beginning of this thread, and it kept giving me error. please help.


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 asia1.ethermine.org:4444 -ewal 0xbF5498Dcb68228c314BE963d4f0913F2dyAdF23E.Miningrig2 -epsw x
 ethpool, ethermine  (and Siacoin pool):
EthDcrMiner64.exe -epool us1.ethpool.org:3333 -ewal 0xD69af2A796A737A103F12d2f0BCC563a1390oE6F.YourWorkerName -epsw x -dpool http://sia-eu1.nanopool.org:9980/miner/header?address=3be0304dee313515cf401b8593a0c1df905ed13f0adaee89a8d7337d2ba8209e5ca9f297bbc2 -dcoin sia

You don't have the correct address for your sia pool.

it should look like this.  

-dpool stratum+tcp://sia-eu1.nanopool.org:7777 -dwal YOUR_SIA_ADDRESS/YOUR_WORKER_NAME/YOUR_EMAIL -dcoin sia

Thanks, um do I exclude the EthDcrMiner? 

[androstan1234]

I cannot even install the driver.  I ran ddu, restarted, then tried to install the driver.  Partway through installation I get a blue screen with stop code "SYSTEM THREAD EXCEPTION NOT HANDLED," and saying that atikmpag.sys failed.

I had the same issue but that was without DDU use, I tried to update the driver by installing it over the other one, after reboot I used DDU and it worked make sure you reboot after each step..

Thanks, I did reboot after each step.  Fixed this by removing all gpus except one.

Getting roughly an extra mhs on each card.

[cdrx]

Claymore 9.8
AMD Driver [Suspicious link removed]p*ute
Gigabyte RX 580 Gaming 8 GB

Voltage doesn't work with Wattman and AB beta 16 or Trixx.

Here is my screen capture,

Using Claymore V9.7 not V9.8

works for you, but I do have 580, not 480. Somehow with "reference design" in AB works, but I am mining eth fork (different dag) - and when dev mining starts and then stops, I am losing my undervoltage.

[token_coin_guy]

updated to 17.7.2 and 9.8 and I cannot keep memory clock at 2045 where I usually keep it.. it keeps bouncing 2045 to 1000 and back to 2045 as a result my hash rate is saw tooth even on my pool page. evn in wattman, the slider is not able to be set to a static value.. seems greyed out.

I installed with DDU to clear old drivers, I also tried upgrading to the new beta drivers and no fix.

it was fine prior to upgrading to 17.7.2

Has anyone else come across this issue?

ASUS ROG Strix Radeon RX 580

https://i.imgur.com/OZ0Zs3C.png

[token_coin_guy]

updated to 17.7.2 and 9.8 and I cannot keep memory clock at 2045 where I usually keep it.. it keeps bouncing 2045 to 1000 and back to 2045 as a result my hash rate is saw tooth even on my pool page. evn in wattman, the slider is not able to be set to a static value.. seems greyed out.

I installed with DDU to clear old drivers, I also tried upgrading to the new beta drivers and no fix.

it was fine prior to upgrading to 17.7.2

Has anyone else come across this issue?

ASUS ROG Strix Radeon RX 580

https://i.imgur.com/OZ0Zs3C.png

disabled secondary monitor and rebooted and I can control memory clock in wattman, but it's still bouncing to 1000 mhz and getting messed up hash

[okiagr]

updated to 17.7.2 and 9.8 and I cannot keep memory clock at 2045 where I usually keep it.. it keeps bouncing 2045 to 1000 and back to 2045 as a result my hash rate is saw tooth even on my pool page. evn in wattman, the slider is not able to be set to a static value.. seems greyed out.

I installed with DDU to clear old drivers, I also tried upgrading to the new beta drivers and no fix.

it was fine prior to upgrading to 17.7.2

Has anyone else come across this issue?

ASUS ROG Strix Radeon RX 580

https://i.imgur.com/OZ0Zs3C.png

disabled secondary monitor and rebooted and I can control memory clock in wattman, but it's still bouncing to 1000 mhz and getting messed up hash

Try to remove the gpu and memory clock settings in your bat file. Leave just the wattman settings.

[token_coin_guy]

updated to 17.7.2 and 9.8 and I cannot keep memory clock at 2045 where I usually keep it.. it keeps bouncing 2045 to 1000 and back to 2045 as a result my hash rate is saw tooth even on my pool page. evn in wattman, the slider is not able to be set to a static value.. seems greyed out.

I installed with DDU to clear old drivers, I also tried upgrading to the new beta drivers and no fix.

it was fine prior to upgrading to 17.7.2

Has anyone else come across this issue?

ASUS ROG Strix Radeon RX 580

https://i.imgur.com/OZ0Zs3C.png

disabled secondary monitor and rebooted and I can control memory clock in wattman, but it's still bouncing to 1000 mhz and getting messed up hash

Try to remove the gpu and memory clock settings in your bat file. Leave just the wattman settings.

and now it's randomly greyed out again in wattman so I can't control it. this is getting crazy. think i am going to roll back driver upgrades

[Jell]

I purchased a RX Vega 64 (standard air cooled) didn't have chance to do anything with it until today. With the standard driver i was getting 31-31Mh/s on Claymore 9.8 ETH only using nanopool.

I have switched to the new AMD Beta driver, i was getting 34Mhs but leaving the fan on auto the temps go through the roof, once it gets near 70 Deg C it would slow down.

Using WattMan, ive underclocked the CPU, undervolted the CPU, overclocked the Memory and undervolted the memory. I have also switched the fan profile to manual, Minimum 75% and Target 100%.

Its been running ETH only for about 3 hours now, im getting steady 43.3 - 43.4 Temp is steady at 63 Deg C

Sadly at the wall my idle usage goes from 70w up to 350w when mining and i wont talk about the fan noise.

[MrJaekin]

I purchased a RX Vega 64 (standard air cooled) didn't have chance to do anything with it until today. With the standard driver i was getting 31-31Mh/s on Claymore 9.8 ETH only using nanopool.

I have switched to the new AMD Beta driver, i was getting 34Mhs but leaving the fan on auto the temps go through the roof, once it gets near 70 Deg C it would slow down.

Using WattMan, ive underclocked the CPU, undervolted the CPU, overclocked the Memory and undervolted the memory. I have also switched the fan profile to manual, Minimum 75% and Target 100%.

Its been running ETH only for about 3 hours now, im getting steady 43.3 - 43.4 Temp is steady at 63 Deg C

Sadly at the wall my idle usage goes from 70w up to 350w when mining and i wont talk about the fan noise.

This is what I was afraid of.. looks like 70-100MH/s was too good to be true. Only time will tell I guess. Once we get past the locked VBIOS hopefully we can get the power usage down and MH/s up.  Don't think we'll be hitting 70-100 though.  Thanks for the update!

[Ursul0]

the amd dag fix driver release seem to be a bit more stable(possibly) version of the custom beta dag fix that was available before.
same lower power consumption and same improved hash numbers...

if you install only drivers (no amd gui shit. try to make a habit of it) you don't even need DDU (which for me causes UEFI win10 setup to get 32bit display driver by default and needs BIOS access two times to boot after ddu).

no need for AB nor Trixx or other crap.
Just tighter straps + undervolt in bios and set clocks and volts in claymore

[Xardas2014]

From what I see so far, the Vega isn't worth the money. I only bought one to test, a water cooled version, but it hasn't arrived yet. From what I'm reading the power usage makes the Vega a terrible choice for mining.

[Tiurog]

Is there a way to get Claymore use the new driver files without installing the package?

I tried to move the new driver file from the beta driver to the claymore folder and EthDcrMiner64.exe are indeed load up the files but the hashrate is still the same old speed from the earlier driver that are installed on Windows. So it looks like Claymore is still using the OS driver files.

files: atig6txx.dll, amdocl12cl64.dll, amd_opencl64.dll->opencl.dll,
I also tested: all files from the driver package into the claymore folder, still no change. (C:\AMD\Win10-64Bit-Crimson-ReLive-Beta-Blockchain-Workloads-Aug11\Packages\Drivers\Display\WT6A_INF\B316900)

@Claymore, is it possible that your miner uses driver files out of your folder? If yes, which one are specific to mining, except the ones i listed.

The problem is, the new drivers are very unstable with under-clocking and only the 17.5.2 driver is the last stable driver that works for me 100%. So i want to keep the old driver while using the new fixed driver for mining.

[joe318]

Hi everybodyyyyy

I have one strange problem, i have one rig with 6 rx series 5, it work well at 2 months ago, but now with this blockchain driver, when i turn off teamviewer he restart, if i stay with teamviewer opened, no problem he mine very well, i win around 9 mh/s.

I have 6 dummies, i know it's a lot, but when i build it, i do a lot tests and i forget take out this dummies lol.

Someone with the same bug?

Thank you

PS i forget told you, now with remote desktop from windows, he work like a charm, before this update .......

[NavyCuda]

It seems as though the blockchain drivers do not support Vega FE.

[Ursul0]

The problem is, the new drivers are very unstable with under-clocking and only the 17.5.2 driver is the last stable driver that works for me 100%. So i want to keep the old driver while using the new fixed driver for mining.

New drivers stability seems to be just fine. you're likely doing something wrong...
not really sure though what exactly, provide more information. like what cards are you using and what are the settings and bioses?

[Ursul0]

Hi everybodyyyyy

I have one strange problem, i have one rig with 6 rx series 5, it work well at 2 months ago, but now with this blockchain driver, when i turn off teamviewer he restart, if i stay with teamviewer opened, no problem he mine very well, i win around 9 mh/s.

I have 6 dummies, i know it's a lot, but when i build it, i do a lot tests and i forget take out this dummies lol.

Someone with the same bug?

Thank you

PS i forget told you, now with remote desktop from windows, he work like a charm, before this update .......

6 dummies is not very smart...
I can recommend RealVNC - works like a charm.

Also dont use RDP, as it seriously fucks with the display drivers, this MS crap is just too high level for it to work proper. (you can seemingly access the running miner, but you cannot start the miner app, as it will see no GPUs. )

Also you can try to use this: https://docs.microsoft.com/en-us/sysinternals/downloads/autologon
(as your issues could be related to running sessions)