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

[Alex1964]

Hey guys, need help, using claymore 9.2,

the mining is okay but there is no hashrate data show on supernova dashboard, anyone can help?

the code i use is -dpool stratum+tcp://dcr.suprnova.cc:2255 -dwal xxxx.test1 -dpsw test

Please see the screenshot below.

https://s9.postimg.org/igv1wzvp7/1111.png
https://s8.postimg.org/k7kp45uf5/22222.png

дoбaвь вoт тaк :

-dpool stratum+tcp://dcr.suprnova.cc:3252 -dwal xxxx.test1 -dpsw test -esm 3 -allpools 1

[christiano88]

I have a problem installing the driver 16.11.4 and after a reboot windows 10 pro changes the driver degrades it to 16.6 as I can solve many thanks

I have exactly the same problem I did not understand.

But it works with 16.6

I don't recommend keeping 16.6, as I know, it has a fan controller bug, which either stops the fans or spins them at 100% randomly (the first one is worse).

[miner582]

For those mining dual +DCR today I notice at coinmine.pl pool that block number is stuck on 127179 so maybe there is another problem with the header update on decred network

[Trimegistus]

I don't recommend keeping 16.6, as I know, it has a fan controller bug, which either stops the fans or spins them at 100% randomly (the first one is worse).

I stopped worrying about drivers versions some time ago. My 6x480 rigs run Win 10 x64 with the 16.6 driver provided by the Microsoft drivers update routine.

It works flawlessly, with Claymore, sg-miner, gateless gate... no fan bug whatsoever.

[heathkun]

I had not much time for testing due to urgent DCR fix, just noticed slowdown for Polaris on Linux. It is because OpenCL reports that it has 14 compute units (CUs) instead of 36. You can fix it by using undocumented option: specify "-eqlim 288". 288 = CUs * 8. I hate this magic, so I'm working on next update that will have good hashrates with 100% stable mining speed on both Windows and Linux. Just wait a few days.

Thanks for the hard work! 

[matmator]

very nice  maj 9.2.

I lose only 5 mh on my 2 rig + pc.

and 4300 mh pascal.

nice job

thx

[matmator]

Where do I find reboot.bash?

No one?

shutdown /r /t 2

[reload88]

Hey guys, need help, using claymore 9.2,

the mining is okay but there is no hashrate data show on supernova dashboard, anyone can help?

the code i use is -dpool stratum+tcp://dcr.suprnova.cc:2255 -dwal xxxx.test1 -dpsw test

Please see the screenshot below.

https://s9.postimg.org/igv1wzvp7/1111.png
https://s8.postimg.org/k7kp45uf5/22222.png

дoбaвь вoт тaк :

-dpool stratum+tcp://dcr.suprnova.cc:3252 -dwal xxxx.test1 -dpsw test -esm 3 -allpools 1

thank you so much, it is working now, btw, what is -esm 3 -allpools 1

dcrpool seems down now?

[notbatman]

I don't recommend keeping 16.6, as I know, it has a fan controller bug, which either stops the fans or spins them at 100% randomly (the first one is worse).

I stopped worrying about drivers versions some time ago. My 6x480 rigs run Win 10 x64 with the 16.5 driver provided by the Microsoft drivers update routine.

It works flawlessly, with Claymore, sg-miner, gateless gate... no fan bug whatsoever.

I've got some MSI cards and the win10 drivers seemed to have chewed up the (just a guess) 32-bit bus running claymore, the cards still run under ubuntu in pure 64bit mode. These cards even have samsung memory and MSI still fucked the dog.

I've got a powercolor and a sapphire running claymore on win10 without any issues and really f'n fast too.

[jackbox]

I don't recommend keeping 16.6, as I know, it has a fan controller bug, which either stops the fans or spins them at 100% randomly (the first one is worse).

I stopped worrying about drivers versions some time ago. My 6x480 rigs run Win 10 x64 with the 16.5 driver provided by the Microsoft drivers update routine.

It works flawlessly, with Claymore, sg-miner, gateless gate... no fan bug whatsoever.

I've got some MSI cards and the win10 drivers seemed to have chewed up the (just a guess) 32-bit bus running claymore, the cards still run under ubuntu in pure 64bit mode. These cards even have samsung memory and MSI still fucked the dog.

I've got a powercolor and a sapphire running claymore on win10 without any issues and really f'n fast too.

I am running an MSI RX 470 Armor 8GB on Windows 10 1703 and it is running very well.

[pornluver]

I noticed that claymore sometimes fail

When it fails it will restart and fail again at the same card

So sometimes I see:

Code:

04:56:49:166	f04	em hbt: 0, fm hbt: 16, 
04:56:49:166	f04	watchdog - thread 0, hb time 250
04:56:49:182	f04	watchdog - thread 1, hb time 78
04:56:49:182	f04	watchdog - thread 2, hb time 79172
04:56:49:182	f04	WATCHDOG: GPU 1 hangs in OpenCL call, exit
04:56:49:182	f04	watchdog - thread 3, hb time 79016
04:56:49:197	f04	WATCHDOG: GPU 1 hangs in OpenCL call, exit
04:56:49:197	f04	watchdog - thread 4, hb time 78
04:56:49:197	f04	watchdog - thread 5, hb time 250
04:56:49:197	f04	watchdog - thread 6, hb time 250
04:56:49:197	f04	watchdog - thread 7, hb time 94
04:56:49:213	f04	watchdog - thread 8, hb time 78
04:56:49:213	f04	watchdog - thread 9, hb time 250
04:56:49:213	f04	watchdog - thread 10, hb time 16
04:56:49:213	f04	watchdog - thread 11, hb time 188
04:56:50:353	f04	Restarting OK, exit...

Then it restart again and I got this

Code:

04:56:54:572	f14	OpenCL platform: AMD Accelerated Parallel Processing
04:56:54:572	f14	OpenCL initializing...

04:56:54:588	f14	AMD Cards available: 6 
04:56:54:588	f14	GPU #0: Ellesmere, 4096 MB available, 36 compute units
04:56:54:588	f14	GPU #0 recognized as Radeon RX 480
04:56:54:588	f14	GPU #1: Ellesmere, 4096 MB available, 36 compute units
04:56:54:588	f14	GPU #1 recognized as Radeon RX 480
04:56:54:604	f14	GPU #2: Ellesmere, 4096 MB available, 36 compute units
04:56:54:604	f14	GPU #2 recognized as Radeon RX 480
04:56:54:604	f14	GPU #3: Ellesmere, 4096 MB available, 36 compute units
04:56:54:604	f14	GPU #3 recognized as Radeon RX 480
04:56:54:619	f14	GPU #4: Ellesmere, 4096 MB available, 36 compute units
04:56:54:619	f14	GPU #4 recognized as Radeon RX 480
04:56:54:619	f14	GPU #5: Ellesmere, 4096 MB available, 36 compute units
04:56:54:619	f14	GPU #5 recognized as Radeon RX 480
04:56:54:635	f14	POOL/SOLO version
04:56:54:635	f14	b255 
04:56:54:635	f14	Platform: Windows
04:56:54:901	f14	start building OpenCL program for GPU 0...
04:57:03:792	f14	done
04:57:03:917	f14	start building OpenCL program for GPU 1...

So definitely GPU 1 is problematic. However, if I wait a few minutes and run it works.

I wonder what's the problem? Is it temperature? Is there a way to log temperature? Is there a way to tell claymore to wait for 1-2 minutes before restarting?

I am not the only one with the problem

https://forum.ethereum.org/discussion/10243/claymore-crashing-watchdog-hang-error

[notbatman]

I don't recommend keeping 16.6, as I know, it has a fan controller bug, which either stops the fans or spins them at 100% randomly (the first one is worse).

I stopped worrying about drivers versions some time ago. My 6x480 rigs run Win 10 x64 with the 16.5 driver provided by the Microsoft drivers update routine.

It works flawlessly, with Claymore, sg-miner, gateless gate... no fan bug whatsoever.

I've got some MSI cards and the win10 drivers seemed to have chewed up the (just a guess) 32-bit bus running claymore, the cards still run under ubuntu in pure 64bit mode. These cards even have samsung memory and MSI still fucked the dog.

I've got a powercolor and a sapphire running claymore on win10 without any issues and really f'n fast too.

I am running an MSI RX 470 Armor 8GB on Windows 10 1703 and it is running very well.

I've got one MSI RX480 Armor card and everything's working OK so far except the fan, it failed after a few weeks.

[notbatman]

@pornluver,

   Get gpu-z and see if you've got Hynix RAM, if so you're fucked and you'll need to RMA. Good luck with that.

[PcChip]

I've got some MSI cards and the win10 drivers seemed to have chewed up the (just a guess) 32-bit bus running claymore, the cards still run under ubuntu in pure 64bit mode. These cards even have samsung memory and MSI still fucked the dog.

this doesn't even make sense.

[Trimegistus]

I've got some MSI cards and the win10 drivers seemed to have chewed up the (just a guess) 32-bit bus running claymore, the cards still run under ubuntu in pure 64bit mode. These cards even have samsung memory and MSI still fucked the dog.

I've got a powercolor and a sapphire running claymore on win10 without any issues and really f'n fast too.

I have 12 Sapphire 480 Nitro+ 4GB, all running stable 24/7 under Win10 with "old" Microsoft drivers

[notbatman]

I've got some MSI cards and the win10 drivers seemed to have chewed up the (just a guess) 32-bit bus running claymore, the cards still run under ubuntu in pure 64bit mode. These cards even have samsung memory and MSI still fucked the dog.

this doesn't even make sense.

Funky artifacts including jagged black triangles on screen when running under windows drivers with claymore accessing the card in 32 bit mode. The card started doing this after running for a week.

vs.

Everything working OK with linux drivers while accessing the card in 64 bit only mode.

[PcChip]

Funky artifacts including jagged black triangles on screen

sounds like VRAM that's unstable  (too low voltage, too high clock, too tight timings, etc)

[notbatman]

Funky artifacts including jagged black triangles on screen

sounds like VRAM that's unstable  (too low voltage, too high clock, too tight timings, etc)

I bet if rip the plate off it says Hynix on those chips. gpu-z is reporting Samsung RAM but it must use a bit of Hynix for something.

OMG stay away from MSI and cards with Hynix RAM or you'll end up so butt-hurt your asshole will drag on the ground behind you as you walk.

[scryptr]

Where do I find reboot.bash?

No one?

shutdown /r /t 2

REBOOT.BASH IS NOT PROVIDED--

In the Claymore directory, do this:

1) Copy start.bash to reboot.bash with "cp start.bash reboot.bash".  This keeps the executable file attributes and permissions.  Depending on your system, you may need to prefix the command "cp" with "sudo".

2) Open "reboot.bash" with nano or vi, and delete the command text.  Again, you may need to use "sudo" to call the editor.

3) Replace the command text with your reboot command and save.  The above suggestion by matmator may work for you, but my system has a reboot script.  I call the reboot script by using the full path to the script.  An example: "/opt/scripts/reboot.sh".  That is the only line in my "reboot.bash".  It works every time Claymore's miner cannot restart on its own.       --scryptr

[pornluver]

Claymore only display GPU temperatur for GPU 0. I want to see or log temperature for all GPU

How to do so?