Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 1104.

CoRpO

full member

Activity: 185

Merit: 100

Quote from: bensam1231 on September 28, 2016, 04:53:14 PM

Quote from: CoRpO on September 28, 2016, 09:54:45 AM

purpose of repost 1 hour later ?

He posted between my original post and when I posted again. I reposted as he more then likely missed the original post in the sea of other posts. I thought it was very pertinent information as someone else mentioned having a similar problem. Having discussed a similar option with Genoil, I think it would be prudent in Claymores miner too.

Although thanks for policing. Your service is invaluable to this thread.

Well I may have been a bit rude, I'm pissed off by the pollution in this thread. Anyway problem solved, wou've been heard Wink

Zedmony

newbie

Activity: 27

Merit: 0

Quote from: citronick on September 28, 2016, 10:20:14 PM

Hi Claymore,

I have a small request to include a field in Ethman, to extract the Crimson version installed on each rig.

If this is useful addition, I hope other members can also support this request so that justification for Claymore to spend the effort to add the field.

Thanks and keep up the good work.

Thanks

Why do you want to extract the Crimson version installed on each rig. What is the extra benefit for doing that?

adamvp

hero member

Activity: 1246

Merit: 708

Quote from: Claymore on September 28, 2016, 03:55:27 PM

Quote from: adamvp on September 28, 2016, 03:37:58 PM

another issue with Linux version, from few days it constantly switching SIA pools every 30-40minutes..
Windows version is ok, even if I revert rigs :S

Code:

SC: http error #28.
SC: Failed to check work on pool, retry in 10 sec..

This error code means "operation timeout".
I can say again and again: don't use Linux if you can use Windows! Most problems are related not to miner but to Linux itself and I have to find workarounds all the time.

Hmm Windows is too expensive, and is unstable if miner crash =windows crash and hard reset is necessary. If on linux fail = just kill the process

citronick

legendary

Activity: 1834

Merit: 1080

---- winter*juvia -----

Hi Claymore,

I have a small request to include a field in Ethman, to extract the Crimson version installed on each rig.

If this is useful addition, I hope other members can also support this request so that justification for Claymore to spend the effort to add the field.

Thanks and keep up the good work.

Thanks

wntrb0y

newbie

Activity: 1

Merit: 0

Having a similar issue here, Linux Mint 17.1, FGLRX version 15.30.3 kernel-3.13.0-37-generic-x86_64, using 3 Radeon R9 Fury X GPUs.

All i get when running the ./start.bash script is :

Code:

????????????????????????????????????????????????????????????????ͻ
? Claymore's Dual ETH + DCR/SC/LBC GPU Miner v7.1 ?
????????????????????????????????????????????????????????????????ͼ

ETH: 7 pools are specified
Main Ethereum pool is us1.ethermine.org:4444
DCR: 0 pool is specified
No protocol specified
No protocol specified
No AMD cards in the list.
No NVIDIA CUDA GPUs detected.
No AMD OPENCL or NVIDIA CUDA GPUs found, exit

But i know for a fact the GPUs are being detected by the system, since the output of

Code:

"aticonfig --lsa"

is :

Code:

* 0. 08:00.0 AMD Radeon (TM) R9 Fury Series
1. 0b:00.0 AMD Radeon (TM) R9 Fury Series
2. 0c:00.0 AMD Radeon (TM) R9 Fury Series

EDIT: v7.1 has been working fine on my 6x Fury X rig with FGLRX v 15.20.3 for 30 hours straight now with no issues.....maybe i should change the driver in my problem rig to this revision?

EDIT 2: Just tried swapping drivers....no luck. Forgot to mention the rig that is working correctly is running on Ubuntu 14.04.5 LTS, and the problem rig that i just swapped drivers on is Mint 17.1 MATE.

Snorremans

newbie

Activity: 1

Merit: 0

i give up and hope someone can help me...

i have 8x nvidia gtx 1080.
windows 7, cause its not yet working under ubuntu 16.04

i have installed cuda but keep getting errors....
i added one image which shows everything i can show or what is needed.
Newest drivers are installed etc etc.
https://i.imgsafe.org/c3a5fb0de8.png

brusque

newbie

Activity: 6

Merit: 0

Quote from: armyman on September 28, 2016, 04:47:46 PM

Which software would you suggest to undervolt/light overclocking? Wattman? Is it possible to download it somewhere?

It comes with AMD drivers. IIRC, you go to AMD Settings | Gaming | Global Settings.

Anyway, I hate those fancy messy gamer apps. WattTool is much faster and easier.

http://www.overclock.net/t/1609782/watttool-a-simple-tool-that-combines-overclocking-with-vrm-monitoring-tweaking-for-rx-400-series

Regards, Brusque

xeridea

sr. member

Activity: 449

Merit: 251

Quote from: armyman on September 28, 2016, 04:47:46 PM

Quote from: xeridea on September 28, 2016, 11:32:18 AM

Quote from: Xardas2014 on September 28, 2016, 05:40:05 AM

armyman your PSU isn't enough. You shouldn't run a PSU at more than 80% of it's rating. rx480's can pull up to about 225 watts each, which is 1,350 watts. You have up to another 150-200 watts for the system also, excluding the vid cards. So you are over 1500 watts on a 1200 watt PSU. Of course it will not be stable. A 1600 watt PSU will run 6 480's and the system....just. It is still pushing the limit a bit much when you factor in the 80% rule....but it will run stable.

480 draws ~165w stock (this is by card, not at the wall). If you don't change power limit, the cards would draw ~ 1000w, and other system parts maybe 50-100, so PSU shouldn't be overloaded, but is higher than you would want for 24x7 operation, and may be slightly less stable near full load. You can undervolt to save ~20-30w per card.

Which software would you suggest to undervolt/light overclocking? Wattman? Is it possible to download it somewhere?

Thanks

Wattman is part of the driver. Just go to Radeon settings, Global Settings. There is a tool someone made called WattTool, I haven't used it though, wattman works just fine.

padrino

legendary

Activity: 1428

Merit: 1000

https://www.bitworks.io

Quote from: armyman on September 28, 2016, 04:39:10 PM

Quote from: Xardas2014 on September 28, 2016, 06:14:00 AM

Quote from: Bakery on September 28, 2016, 05:55:43 AM

Quote from: Xardas2014 on September 28, 2016, 05:40:05 AM

armyman your PSU isn't enough. You shouldn't run a PSU at more than 80% of it's rating. rx480's can pull up to about 225 watts each, which is 1,350 watts. You have up to another 150-200 watts for the system also, excluding the vid cards. So you are over 1500 watts on a 1200 watt PSU. Of course it will not be stable. A 1600 watt PSU will run 6 480's and the system....just. It is still pushing the limit a bit much when you factor in the 80% rule....but it will run stable.

The miner doesnt start - this isnt a problem by psu power. If psu power fails, the system shuts down.

It is the PSU, during start up and DAG process the system draws more power. His ETH process is starting, but stops. Easy way to find out.....disable all the cards but one in the command line options, then enable them one at a time until the crashing starts again Wink

I have tried as you said and it did work!

Unfortunately I have just been able to modify the setup from my phone through remote control, but apparently with one card only it was starting mining....I have noticed that it the beginning the hashing speed was 0, is that normal?

Yes that is normal, until the DAGs finish..

bensam1231

legendary

Activity: 1750

Merit: 1024

Quote from: Claymore on September 28, 2016, 10:30:43 AM

Quote from: bensam1231 on September 27, 2016, 08:22:40 PM

I am going to mirror the sentiment that intensity needs to be reduced during dag generation. Previous memory OCs were stable, now with the latest epoc they are no longer stable during generation. Please add the ability to reduce intensity on the miner during DAG generation.

I can generate the dag without a memory OC just fine then OC the memory after the miner is running and it works. If I try to generate the DAG while the memory OC is in place the GPUs crash.

Ok, in next update I'll add "-lidag" option. New version will be available in 1-2 days.

Appreciate it dude.

Quote from: CoRpO on September 28, 2016, 09:54:45 AM

purpose of repost 1 hour later ?

He posted between my original post and when I posted again. I reposted as he more then likely missed the original post in the sea of other posts. I thought it was very pertinent information as someone else mentioned having a similar problem. Having discussed a similar option with Genoil, I think it would be prudent in Claymores miner too.

Although thanks for policing. Your service is invaluable to this thread.

Quote from: Claymore on September 28, 2016, 03:55:27 PM

Quote from: adamvp on September 28, 2016, 03:37:58 PM

another issue with Linux version, from few days it constantly switching SIA pools every 30-40minutes..
Windows version is ok, even if I revert rigs :S

Code:

SC: http error #28.
SC: Failed to check work on pool, retry in 10 sec..

This error code means "operation timeout".
I can say again and again: don't use Linux if you can use Windows! Most problems are related not to miner but to Linux itself and I have to find workarounds all the time.

Weird to see a developer who isn't singing praises about Nix and wearing his Nix badge of honor around. Almost like there is more to the OS then just one is better then the other.

armyman

newbie

Activity: 45

Merit: 0

Quote from: xeridea on September 28, 2016, 11:32:18 AM

Quote from: Xardas2014 on September 28, 2016, 05:40:05 AM

armyman your PSU isn't enough. You shouldn't run a PSU at more than 80% of it's rating. rx480's can pull up to about 225 watts each, which is 1,350 watts. You have up to another 150-200 watts for the system also, excluding the vid cards. So you are over 1500 watts on a 1200 watt PSU. Of course it will not be stable. A 1600 watt PSU will run 6 480's and the system....just. It is still pushing the limit a bit much when you factor in the 80% rule....but it will run stable.

480 draws ~165w stock (this is by card, not at the wall). If you don't change power limit, the cards would draw ~ 1000w, and other system parts maybe 50-100, so PSU shouldn't be overloaded, but is higher than you would want for 24x7 operation, and may be slightly less stable near full load. You can undervolt to save ~20-30w per card.

Which software would you suggest to undervolt/light overclocking? Wattman? Is it possible to download it somewhere?

Thanks

armyman

newbie

Activity: 45

Merit: 0

Quote from: Xardas2014 on September 28, 2016, 06:14:00 AM

Quote from: Bakery on September 28, 2016, 05:55:43 AM

Quote from: Xardas2014 on September 28, 2016, 05:40:05 AM

armyman your PSU isn't enough. You shouldn't run a PSU at more than 80% of it's rating. rx480's can pull up to about 225 watts each, which is 1,350 watts. You have up to another 150-200 watts for the system also, excluding the vid cards. So you are over 1500 watts on a 1200 watt PSU. Of course it will not be stable. A 1600 watt PSU will run 6 480's and the system....just. It is still pushing the limit a bit much when you factor in the 80% rule....but it will run stable.

The miner doesnt start - this isnt a problem by psu power. If psu power fails, the system shuts down.

It is the PSU, during start up and DAG process the system draws more power. His ETH process is starting, but stops. Easy way to find out.....disable all the cards but one in the command line options, then enable them one at a time until the crashing starts again Wink

I have tried as you said and it did work!

Unfortunately I have just been able to modify the setup from my phone through remote control, but apparently with one card only it was starting mining....I have noticed that it the beginning the hashing speed was 0, is that normal?

Claymore

donator

Activity: 1610

Merit: 1325

Miners developer

Quote from: adamvp on September 28, 2016, 03:37:58 PM

another issue with Linux version, from few days it constantly switching SIA pools every 30-40minutes..
Windows version is ok, even if I revert rigs :S

Code:

SC: http error #28.
SC: Failed to check work on pool, retry in 10 sec..

This error code means "operation timeout".
I can say again and again: don't use Linux if you can use Windows! Most problems are related not to miner but to Linux itself and I have to find workarounds all the time.

adamvp

hero member

Activity: 1246

Merit: 708

another issue with Linux version, from few days it constantly switching SIA pools every 30-40minutes..
Windows version is ok, even if I revert rigs :S

Code:

SC: http error #28.
SC: Failed to check work on pool, retry in 10 sec..

fredeq

legendary

Activity: 1537

Merit: 1005

Quote from: adamvp on September 28, 2016, 01:48:51 PM

Quote from: Claymore on September 28, 2016, 12:01:54 PM

Quote from: adamvp on September 28, 2016, 11:13:49 AM

bad issue with SSH linux strikes again Sad

To remind: When I want to end/break/kill screen session miner don't want to hear me Tongue

Today I noticed that then GPU hangs and miner restart info..
In previous versions everything was OK...

How can I reproduce this issue?

I don't know if it was any backend changes, but V 7.1 introduce colors. There is no big problem for me now, because I have only one linux rig, but if I will make more in the future.. or you could add "on the run" pool management in any future realease Wink

Oh yes, live pool switch please Cheesy

adamvp

hero member

Activity: 1246

Merit: 708

Quote from: Claymore on September 28, 2016, 12:01:54 PM

Quote from: adamvp on September 28, 2016, 11:13:49 AM

bad issue with SSH linux strikes again Sad

To remind: When I want to end/break/kill screen session miner don't want to hear me Tongue

Today I noticed that then GPU hangs and miner restart info..
In previous versions everything was OK...

How can I reproduce this issue?

I don't know if it was any backend changes, but V 7.1 introduce colors. There is no big problem for me now, because I have only one linux rig, but if I will make more in the future.. or you could add "on the run" pool management in any future realease Wink

citronick

legendary

Activity: 1834

Merit: 1080

---- winter*juvia -----

Quote from: Claymore on September 28, 2016, 12:05:38 PM

Quote from: Nippy-ua on September 28, 2016, 11:38:37 AM

stuck...

It seems now AMD drivers have problems because OpenCL does not work, miner asks OpenCL about AMD platform and it returns nothing. So the reason is in drivers.

Claymore,

Will it help if Ubuntu 14 instead of 16.

Seems v16 doesnt have the complete AMD drivers.

Nippy-ua

newbie

Activity: 48

Merit: 0

Quote from: Claymore on September 28, 2016, 12:05:38 PM

Quote from: Nippy-ua on September 28, 2016, 11:38:37 AM

stuck...

It seems now AMD drivers have problems because OpenCL does not work, miner asks OpenCL about AMD platform and it returns nothing. So the reason is in drivers.

yep
I'm trying to install a new system
thanks

Claymore

donator

Activity: 1610

Merit: 1325

Miners developer

Quote from: Nippy-ua on September 28, 2016, 11:38:37 AM

stuck...

It seems now AMD drivers have problems because OpenCL does not work, miner asks OpenCL about AMD platform and it returns nothing. So the reason is in drivers.

citronick

legendary

Activity: 1834

Merit: 1080

---- winter*juvia -----

Quote from: Nippy-ua on September 28, 2016, 11:38:37 AM

stuck...

I thought Claymore said that he would fix the limit in next release.... so your 19 x GPUs may not work yet with current release.

You should take some pictures to share with us - its an amazing project

Topic: Claymore's Dual Ethereum AMD+NVIDIA GPU Miner v15.0 (Windows/Linux) - page 1104. (Read 6590565 times)