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

[black1ight]

CLAYMORE STUCK AFTER START

Who is the magic wizard that can help me out,

I have been going to all posts, but I cant get a suitable solution for my problem with Claymore 12.0. Tried everything.


Later versions of claymore give me the Can not create DAG file, problem.  Which is said to be a virt mem problem. (i ve put the virt mem to even 50GB, and even tried 80GB) went throught the CHKDSK /f /r solutions and sata AHCI controller driver solutions.
Or I get a video TDR failure, which should be a driver issue.


I cant figure out why half of my miners suddenly are stuck... and the rest are working fine... ? exactelly same BIOS setup, same CPU, same DDR. Both miners are in every way: an exact clone.

Throuble shooting:
swapped all risers with GPUs from the working one to the one that stucks.
i went through BIOS, 4G is on. tried GEN1 and GEN2....GEN AUTO   (the working miner is actually even on GEN  AUTO)
But...
So when i swap the SSD only from the working RIG to the exact same RIG, that isnt working... claymore gets stuck again....

Tried to restart / reload windows 6 times.
With first having 1 GPU, and adding everytime an extra one.

I have a 6 GPU 1080TI card, with the latest Nvidia drivers installed, (used DDU, took the Latest Nvidea driver, and as well tried 425.1 and 430.86)  
Asrock BTC pro 2.0 board latest Bios update with the 4g decoding option which is ON) , with 16GB DDR, intel i5 4core processor. 256 samsung EVO SSD.


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 us1.ethermine.org:4444 -ewal 0xc20ca97592ecfbd89a86df556b7784376a672e30.miner1 -epsw x -dpool stratum+tcp://lbry.suprnova.cc:6256 -dwal my.miner1 1 -dpsw x -dcoin lbc


what on earth can it be? what am i over seeing wrong for the last weeks?
wont make sence why same hardware works with one miner and not with the other one. I tried swappin SSD, or cloning an IMG file.


additional info:
Thanks a lot for the reply,

I am Running on windows
I have not (yet) any afterburner installed. i am not overclocking and have the full power (1600W ) available


Claymore 12.0
When i run claymore 12.0 , its finds the pool and after create DAG for each GPU, the screen freezes, and i have to manually (swich of power) restart
when i wait longer i get the blue windows screen DPC watchdog violation

Claymore 14.7
When i run claymore 14.7 , its is running bit better, One of the GPU s (sometime GPU4 and somethimes GPU 3) is on 0
NVML: can not get current temperature error 15.
then it reboots with one GPU less.



Any help Would be Appreciated, and rewarded from my side!!


thanks in advance

Hi,

Had a similar issue.

For me the issue was due to the power cable to the riser. .

[Metroid]

@claymore, anyway you can disclose your timings if we decide to use on our own?

[boregepco]

Core Clock is up and down - why it can't stay on one place?

[fapar]

Core Clock is up and down - why it can't stay on one place?

How often? This can be when reconnecting to pools (your pool and dev-fee pool), the formation of a new DAG's. If it's remote farm while on the Teamviewer may also fall frequency.

[boregepco]

I am making everything on my pc, without a Teamviewer.
I want Core Clock to be 1150, but it is always somewhere - up or down.
This starts may be 2-3  min from the beginning, it is not connected with a poll change.
May be GPU temperature is the problem - when the temp is around 71-72 C...
But I haven't tt or other parameters about temp in my config...

[fapar]

I am making everything on my pc, without a Teamviewer.
I want Core Clock to be 1150, but it is always somewhere - up or down.
This starts may be 2-3  min from the beginning, it is not connected with a poll change.
May be GPU temperature is the problem - when the temp is around 71-72 C...
But I haven't tt or other parameters about temp in my config...

First, show the startup parameters of the Claymore;
Second, install any monitoring programme (HWiNFO64 or OpenHardwareMonitor).
Most likely your problem is related to the set temperature limit parameters. For example, there is a temperature limit of 70 (-tt 70) and when the card is heated above this temperature it resets the frequency.

[keksik]

hello,

i have latest wn 10, amd 19.5.2 and claymore 14.7 and despsite i use rxboost or strapo value, still getting same ahsrate with or without it.
i had 14.6, downloaded 14.7, used driver uninstall, rebooted, then driver install, again reboot but still getting same hash.

it wrote rxoption is applied succesfuly, same for strap, but i do not see any changes in hashrate...any advice what could be wrong?

[greasepit]

I got some improve on my hash rate speed. Before it usually used to work at 177Mh/s and now 188Mh/s.
I did the same config for the 3 rigs I have and the total gain was 30Mh/s it is like I have gotten another GPU  

However I'm gonna monitor my graph on the pool and see if I really had some gain... Let's see!!

https://i.ibb.co/9WY90LR/cm147.png


- 17 Cards are Asus Radeon RX 580 OC 4GB - samsung memory
- 1 Card is Asus Radeon RX 580 OC 4GB - elpida memory

[Yurius_R]

I have some test rig on windows machine with some different cards and with latest 14.7 I saw a nice new feature - showing power consumption of every card and total of the rig. But for some reason there is no info on power consumption on my R9 380 (GPU2 on log). Is there any way to fix it by me or it's just a bug?

Code:

GPU0 t=63C fan=87% P=52W, GPU1 t=62C fan=56% P=81W, GPU2 t=59C fan=33%, GPU3 t=62C fan=62% P=75W, GPU4 t=55C fan=60% P=74W
Total GPUs power consumption: 282 Watts

Adrenalin 19.4.3
Windows 10 LTSC 1809

[WhackOBill]

hello,

i have latest wn 10, amd 19.5.2 and claymore 14.7 and despsite i use rxboost or strapo value, still getting same ahsrate with or without it.
i had 14.6, downloaded 14.7, used driver uninstall, rebooted, then driver install, again reboot but still getting same hash.

it wrote rxoption is applied succesfuly, same for strap, but i do not see any changes in hashrate...any advice what could be wrong?

The latest Win10 has a new feature called Tamper Protection.  It seems to prevent writing directly to the card's I2C interfaces even with the proper driver.  That may be the problem.  I found that it would prevent AMDMemTweak from adjusting the values as well.  You can turn it off but I can't recall how.  It gave me the option at one point and I took it.

Might not be your problem, but worth a look.  I've been bouncing around with various versions of the AMD driver as well trying to get around this.  I found that 19.6.2 would crash on my machine when going to the settings page.  I backed off to 19.5.2 and it works ok.  I'm only running Claymore 11.9 though.  Waiting for the 14.x to stabilize.

[RuMiner]

Hello! Is there a way to connect to pools via external proxy/socks?

[keksik]

hello,

i have latest wn 10, amd 19.5.2 and claymore 14.7 and despsite i use rxboost or strapo value, still getting same ahsrate with or without it.
i had 14.6, downloaded 14.7, used driver uninstall, rebooted, then driver install, again reboot but still getting same hash.

it wrote rxoption is applied succesfuly, same for strap, but i do not see any changes in hashrate...any advice what could be wrong?

The latest Win10 has a new feature called Tamper Protection.  It seems to prevent writing directly to the card's I2C interfaces even with the proper driver.  That may be the problem.  I found that it would prevent AMDMemTweak from adjusting the values as well.  You can turn it off but I can't recall how.  It gave me the option at one point and I took it.

Might not be your problem, but worth a look.  I've been bouncing around with various versions of the AMD driver as well trying to get around this.  I found that 19.6.2 would crash on my machine when going to the settings page.  I backed off to 19.5.2 and it works ok.  I'm only running Claymore 11.9 though.  Waiting for the 14.x to stabilize.

well i have disable windows defender in group policies... tamper protection is disabled, however i can see realtime protection is still on...so will try tomorrow disable realtime protection and will see how it will go.

EDIT: no chance after disabling realtime protection...

[stef_stef]

What could be the reason for this message?

"opencl error -4 cannot write buffer DAG" after that miner freezes and I have to restart the rig.


I have virtual memory set to 20+ gigs on 8 GPU rig

[jeremyismer]

What could be the reason for this message?

"opencl error -4 cannot write buffer DAG" after that miner freezes and I have to restart the rig.


I have virtual memory set to 20+ gigs on 8 GPU rig

Check your risers. I had that recently, I unplugged, gave them a blast out and away I went.

[RuMiner]

What could be the reason for this message?

"opencl error -4 cannot write buffer DAG" after that miner freezes and I have to restart the rig.


I have virtual memory set to 20+ gigs on 8 GPU rig

each card requires DAG size of memory (3+ Gb for ETH now) + some memory for Windows. you need up to 30 Gb swap

[keksik]

What could be the reason for this message?

"opencl error -4 cannot write buffer DAG" after that miner freezes and I have to restart the rig.


I have virtual memory set to 20+ gigs on 8 GPU rig

use virtual file 8gpu x gpu ram size, for example 8x 4 = 32gb if u have 4gb cards. if you have 8gb, then 64gb cards and u are safe for future dag size+ some miners are option to use double size.

[angelbbs]

v14.7:

- now the miner driver does not require switching Windows to Test Mode. If you want to turn Test Mode off, use "-driver uninstall" option (with admin rights) and reboot,
   then start miner with admin rights to install new driver automatically when necessary (or use "-driver install" option directly).
   If you have any problems with signed driver you can use "-driver install_test" option, it uses old approach: enables Test Mode (reboot is required) and installs unsigned driver.
- added "-showpower" option for stats about GPU power consumption, press "s" to see it. Both AMD and Nvidia cards are supported, except Linux gpu-pro drivers and Radeon7 cards.
- updated Remote Manager to show total power consumption.
- fixed issue with "-logsmaxsize" option when "-logfile" option is used to specify directory for log files.
- a few minor bug fixes and improvements.

Please add DAG epoch number in API responce.

[stef_stef]

What could be the reason for this message?

"opencl error -4 cannot write buffer DAG" after that miner freezes and I have to restart the rig.


I have virtual memory set to 20+ gigs on 8 GPU rig

use virtual file 8gpu x gpu ram size, for example 8x 4 = 32gb if u have 4gb cards. if you have 8gb, then 64gb cards and u are safe for future dag size+ some miners are option to use double size.

I had around 32 gigs actually, and now increased it to 38gigs. I no longer get the error message, however the miner crashes after a minute of mining, and nothing is written in the log file. I am mining with the default clock values and 14.7 miner

[UselessGuru]

v14.7:

- now the miner driver does not require switching Windows to Test Mode. If you want to turn Test Mode off, use "-driver uninstall" option (with admin rights) and reboot,
   then start miner with admin rights to install new driver automatically when necessary (or use "-driver install" option directly).
   If you have any problems with signed driver you can use "-driver install_test" option, it uses old approach: enables Test Mode (reboot is required) and installs unsigned driver.
- added "-showpower" option for stats about GPU power consumption, press "s" to see it. Both AMD and Nvidia cards are supported, except Linux gpu-pro drivers and Radeon7 cards.
- updated Remote Manager to show total power consumption.
- fixed issue with "-logsmaxsize" option when "-logfile" option is used to specify directory for log files.
- a few minor bug fixes and improvements.

Please add DAG epoch number in API responce.

Yes. Please!!!

And add best intensities found in auto tuning mode.

[stef_stef]

What could be the reason for this message?

"opencl error -4 cannot write buffer DAG" after that miner freezes and I have to restart the rig.


I have virtual memory set to 20+ gigs on 8 GPU rig

use virtual file 8gpu x gpu ram size, for example 8x 4 = 32gb if u have 4gb cards. if you have 8gb, then 64gb cards and u are safe for future dag size+ some miners are option to use double size.

I had around 32 gigs actually, and now increased it to 38gigs. I no longer get the error message, however the miner crashes after a minute of mining, and nothing is written in the log file. I am mining with the default clock values and 14.7 miner

It just crashed with one GPU mining and 42 gigs. Any troubleshooting tips?