[OS] nvOC easy-to-use Linux Nvidia Mining

[fullzero]

fullzero sorry i haven't pulled the trigger on this i really do want to but windows is just so damn easy to set up and use.

Hate it as you will - 99% of the influx into the mining scene will be people that can't find their ass from the keyboard, let alone know what Linux is - so I'm just throwing some thing out there for you to consider

Since you are marketing this as "easy-to-use Linux Nvidia mining" here are some things that KEEP me on M$ crappy Win10

Ease of use -

• Windows - obviously it's very familiar and easy to use for the layman
  MSI Afterburner - easy to control GPU's
  .bat files easy to alter
  Remoting in extremely easy - TeamViewer

If someone wants to use windows to mine; that is fine.  As a long time miner I wouldn't recommend it; but whatever floats your boat.

However most of your reasoning is; misinformed.

First; Ubuntu 16.04 is GUI based and easy to use.  With nvOC it takes considerably less time to configure and start mining than it does using windows.

Second; I think it is easier to edit OC values in a bash file myself: but if you are so inclined you can also use the nvidia gui interface (very similar to AB ) with nvOC.

Third; a bat file is derivative of a bash script; it is essentially the windows version of a bash script; I don't understand how that is supposed to be different.

Finally; You can install teamviewer on nvOC; several members have already done this.

PS: you can install windows 10 without a key; this is not "pulling one over on Microsoft": they simply use more of your computers resources for their botnet when you don't have a key.  You are essentially turning your computer into a Microsoft server, when you install windows 10 without a key.  

[1714001481]

1714001481

[1714001481]

1714001481

[1714001481]

1714001481

[1714001481]

1714001481

[1714001481]

1714001481

[UberDaemon]

$250 is more than I want to spend on a tent; but if I can find a used one, I'll probably get it and try it out.  I have used plastic to control airflow before and it was effective.

Good to know stable EVGA 1070 settings for ETH.

If I run across a good deal on one I'll PM ya.  

Regarding EVGA 1070 FTW settings: I may have spoken prematurely.  I did experience a claymore reset overnight.  I will be trying an MC of 1050 and then 1000 if necessary to see if I can get it to run without anymore Claymore drops.  At least when it does drop it immediately resumes, so I'm only losing ~45 secs of hashing, but I know from a long term standpoint it's probably unwise not to address it, so I will dial it back just a bit to pursue 100% error-free operation.

I received my 14th card from Fedex this morning, so now both rigs are running 7 GPUs.  The newest addition is an MSI 1070 Armor.  It's been hashing along for 45m now along with it's EVGA 1070 FTW & SC cousins at the same -200/+1100/125W settings they use so off to a good start.  I do notice the MSI runs about 7-8C hotter than the EVGAs do (they average around 51C at 70% fan speed).

Also my HDMI dummies showed up today, so I can shut off these monitors and save a few watts of power now.  

[fullzero]

Same happens to my on putty, 9x gtx 1060 3gb

Try the procedure from this post and tell me if it solves the problem:

https://bitcointalk.org/index.php?topic=1854250.msg19449945#msg19449945

Also, if you are connecting a monitor directly to the motherboard; connect it to the primary GPU instead (one connected to the first 16x slot closest to the CPU).

[fullzero]

Hey fullzero,

The other issues I've had were resolved automagically - no idea what happened.

Anyhow, I'm encountering a new issue when starting the server:

-I get stuck at the Ubuntu login prompt. m1 is selected as a user and I enter miner1 as a password - I don't get logged in, the screen refreshes and I get put back into the login screen.

Any idea what might be causing this? I'd like to prevent this particular issue from happening as I'll be deploying nvOC to multiple servers and getting everything to work automatically would be awesome.

Edit: I have already tried this - https://bitcointalk.org/index.php?topic=1854250.msg18789262#msg18789262
Edit 1: Tried this as well - http://www.linuxslaves.com/2016/05/3-ways-fix-ubuntu-gets-stuck-login-loop.html = no luck
Edit 2: Will try a fresh install again. It might be a one-time thing

Do the servers you plan on using have more than 1x GPU?

[fullzero]

I did find some OC numbers my EVGA 1070s like and run stable at: -200/1100, ~29.1 MH/sec ETH.  

Thanks again for everything.  Really looking forward to v0016!  

I'm having trouble OC my cards. I have 4x EVGA 1070 SC. Am I doing this right?

I could not get it working by setting INDIVIDUAL_CLOCKS="NO" and I set the individual clocks and -200/1100 to 2 GPUs to see if it worked. I'm still getting 26MH/s.

Note: I'm not sure how to check memory type in Linux. I might have Micron memory.




I can get it working on my 1 card W7 rig:

Are you using a single GPU?

if so look at this post:

https://bitcointalk.org/index.php?topic=1854250.msg19449945#msg19449945

let me know if this is the case / if this fix works

[UberDaemon]

Hey fullzero,

The other issues I've had were resolved automagically - no idea what happened.

Anyhow, I'm encountering a new issue when starting the server:

-I get stuck at the Ubuntu login prompt. m1 is selected as a user and I enter miner1 as a password - I don't get logged in, the screen refreshes and I get put back into the login screen.

Any idea what might be causing this? I'd like to prevent this particular issue from happening as I'll be deploying nvOC to multiple servers and getting everything to work automatically would be awesome.

Edit: I have already tried this - https://bitcointalk.org/index.php?topic=1854250.msg18789262#msg18789262
Edit 1: Tried this as well - http://www.linuxslaves.com/2016/05/3-ways-fix-ubuntu-gets-stuck-login-loop.html = no luck
Edit 2: Will try a fresh install again. It might be a one-time thing

Do the servers you plan on using have more than 1x GPU?

I have experienced that exact same issue and it was with 7x gpu (the 7x 1070  reference/founders rig).  I threw in a freshly imaged drive and it has not reoccurred.

[peonv]

Lxde would be a much lighter GUI than Unity.

[fullzero]

Does setting MEMORY_OVERCLOCK=800 like setting 'MEM CLOCK OFFSET=+800? Do you have any reading material on this? I cannot find anything to help me.

When using a positive offset OC:  DO NOT use the + sign:

However; when using a negative offset OC:  YOU MUST use the - sign.

[UberDaemon]

Lxde would be a much lighter GUI than Unity.

I will 2nd the props for lxde, I have used it on debian and ubuntu EC2 cloud builds (video CDN) and can attest that it is very light on resource usage.  It would likely be an excellent base for nvOC.

It's also fairly easy to use xrdp w/ lxde to run vnc over rdp so wind0ze users can connect via rdp and not have to install any 3rd party apps for remote access.  https://comtechies.com/how-to-set-up-gui-on-amazon-ec2-ubuntu-server.html

[fullzero]

@dev.

Your zip file is unzippable with ubuntu (error message), only with Windows..

Via wich command in your script can you made +500 MH at the memory clock : nvidia-smi don't accept a so big value (I do'nt ask how the user of your script do that, it is the the conf file, but how you have done this, with wich command ?).

Interesting: I will test unzipping with linux.

How much you can set the OC offset is dependent on the specific card / cards you are using.  If you are using multiple different cards you can only use OC values that are valid for all of them when using the general core and memory OC.

If you want to find the limit for a specific card, you can enter a manual OC command for the card in the guake terminal (that you know to be invalid) you will then get a message telling you the valid range for clocks on that card.

You must use individual card clocks to be able to fully OC different cards on the same rig.

[fullzero]

Anyone to let me know what could be the possible cause of nvOC not booting up?

Often times I get a black screen and it will stay that way for 15+min. So far, the only possible solution to this problem is to reinstall nvOC on my usb


MOBO asus prime z270p
g3930
nvidia 1070

By any chance did you:

forget to unzip the image before you imaged the USB?

[fullzero]

Lxde would be a much lighter GUI than Unity.

This is true.

If I end up making a lightweight version eventually; I will build it with arch:

https://www.archlinux.org/

[fullzero]

Lxde would be a much lighter GUI than Unity.

I will 2nd the props for lxde, I have used it on debian and ubuntu EC2 cloud builds (video CDN) and can attest that it is very light on resource usage.  It would likely be an excellent base for nvOC.

It's also fairly easy to use xrdp w/ lxde to run vnc over rdp so wind0ze users can connect via rdp and not have to install any 3rd party apps for remote access.  https://comtechies.com/how-to-set-up-gui-on-amazon-ec2-ubuntu-server.html

I think the Ubuntu RDP application also works with windows, but I will have to test it.

[Nexillus]

Anyone have any suggestions to this problem, I am all ears.

I have one of my rigs which uses a Asus Prim z270-a when I put the 6th GPU in it will not POST to the bios screen therefore not start up.

From troubleshooting it only happens if I have 2 GPUs plugged into the X16 slots, if I only have one it works fine.

I have changed the BIOS to the recommend from the thread here for the z270-a and set it all to GEN2.

Any other suggestions?

[WarwickNZ]

Anyone have any suggestions to this problem, I am all ears.

I have one of my rigs which uses a Asus Prim z270-a when I put the 6th GPU in it will not POST to the bios screen therefore not start up.

From troubleshooting it only happens if I have 2 GPUs plugged into the X16 slots, if I only have one it works fine.

I have changed the BIOS to the recommend from the thread here for the z270-a and set it all to GEN2.

Any other suggestions?

There were 4 Gen2 settings I needed to change, did you change them all? Could be an issue running 2 on the board with 7 or more, i've only tested with 1 on the board +6 risers or all risers. Check for faulty riser cables too, that slowed up a friend of mine getting 7 going on this board.

[weareken]

First, thank you so much fullzero  for this.  I've rolled my own in the past, so this was a nice find and saved me a lot of trouble to do my own.

Have a couple of questions
1) I'm using 4 x PNY GTX 1060 3GB cards on a Gigabyte Z97 SOC Force motherboard.  Under Win10 with power 70, cclock 150, and mclock 850 I consistently get ~23 Mh/s.  With nvOC with the same settings I get 21.4 Mh/s.  Increasing the mclock doesn't seem to increase the hashrate.  cclock at 100 doesn't seem to make a difference either.

My oneBash looks like this:

Code:

COIN="ETH"

POWERLIMIT="70"                 # YES NO

INDIVIDUAL_POWERLIMIT="NO"      # YES NO

__CORE_OVERCLOCK=150
MEMORY_OVERCLOCK=850

INDIVIDUAL_CLOCKS="NO"          # YES NO

When the bash file starts it shows that the attributes are set, but it appears that the clock settings aren't taking effect. I don't see any errors thrown when Claymore starts up except that the cuda library doesn't have a version number.

2) I added a couple of parameters to the Eth command line, but it doesn't seem to pick them up.  For instance, when try to interactively change the dcri it tells me that i need to set -asm 1, but that is already set in the bash file.

Code:

if [ $COIN == "ETH" ]
then
HCD='/home/m1/9.0/ethdcrminer64'
ETHADDR="$ETH_ADDRESS/$ETH_WORKER"

if [ $ETHERMINEdotORG == "YES" ]
then
ETHADDR="$ETH_ADDRESS.$ETH_WORKER"
fi

until $HCD -epool $ETH_POOL -ewal $ETHADDR -epsw x -mode 1 -esm 0 -estale 0 -asm 1 -dcri 12
   do
   echo "FAILURE; reinit in 5" >&2
   sleep 5
done
fi

Any recommendations would be much appreciated.  Windows 10 sucks for mining - so much babysitting, so hoping to get this tweaked and running well.

[fullzero]

Anyone have any suggestions to this problem, I am all ears.

I have one of my rigs which uses a Asus Prim z270-a when I put the 6th GPU in it will not POST to the bios screen therefore not start up.

From troubleshooting it only happens if I have 2 GPUs plugged into the X16 slots, if I only have one it works fine.

I have changed the BIOS to the recommend from the thread here for the z270-a and set it all to GEN2.

Any other suggestions?

There were 4 Gen2 settings I needed to change, did you change them all? Could be an issue running 2 on the board with 7 or more, i've only tested with 1 on the board +6 risers or all risers. Check for faulty riser cables too, that slowed up a friend of mine getting 7 going on this board.

This was my thought as well; WarwickNZ.

I'm not sure if you can use 2x GPUs direct on the mobo + 5x with risers with this mobo.

When using a GPU directly more pcie lanes / pcie bandwidth is used.  This may result in insufficient bandwidth on one of the cards to enable normal operation.

While using a card directly I would change the pcie bandwidth to gen2 for all cards.  Then when you get your risers, I would change it back to auto. 

As WarwickNZ indicated; there are multiple pcie bandwidth related settings in the bios.  There may also be an SLI setting that needs to be disabled as well.

[Nexillus]

Anyone have any suggestions to this problem, I am all ears.

I have one of my rigs which uses a Asus Prim z270-a when I put the 6th GPU in it will not POST to the bios screen therefore not start up.

From troubleshooting it only happens if I have 2 GPUs plugged into the X16 slots, if I only have one it works fine.

I have changed the BIOS to the recommend from the thread here for the z270-a and set it all to GEN2.

Any other suggestions?

There were 4 Gen2 settings I needed to change, did you change them all? Could be an issue running 2 on the board with 7 or more, i've only tested with 1 on the board +6 risers or all risers. Check for faulty riser cables too, that slowed up a friend of mine getting 7 going on this board.

This was my thought as well; WarwickNZ.

I'm not sure if you can use 2x GPUs direct on the mobo + 5x with risers with this mobo.

When using a GPU directly more pcie lanes / pcie bandwidth is used.  This may result in insufficient bandwidth on one of the cards to enable normal operation.

While using a card directly I would change the pcie bandwidth to gen2 for all cards.  Then when you get your risers, I would change it back to auto. 

As WarwickNZ indicated; there are multiple pcie bandwidth related settings in the bios.  There may also be an SLI setting that needs to be disabled as well.

I should probably rephrase. I have all 6 GPUs on risers, the only time it does not post is when I have 2 of the x16 slots with the risers in them. I will double check and make sure they are all the same gen type. I had it on auto first but when I did, it wouldn't even boot past one card. I'm to the point I may just reset the MB bios via cmos and start fresh.

[fullzero]

First, thank you so much fullzero  for this.  I've rolled my own in the past, so this was a nice find and saved me a lot of trouble to do my own.

Have a couple of questions
1) I'm using 4 x PNY GTX 1060 3GB cards on a Gigabyte Z97 SOC Force motherboard.  Under Win10 with power 70, cclock 150, and mclock 850 I consistently get ~23 Mh/s.  With nvOC with the same settings I get 21.4 Mh/s.  Increasing the mclock doesn't seem to increase the hashrate.  cclock at 100 doesn't seem to make a difference either.

My oneBash looks like this:

Code:

COIN="ETH"

POWERLIMIT="70"                 # YES NO

INDIVIDUAL_POWERLIMIT="NO"      # YES NO

__CORE_OVERCLOCK=150
MEMORY_OVERCLOCK=850

INDIVIDUAL_CLOCKS="NO"          # YES NO

When the bash file starts it shows that the attributes are set, but it appears that the clock settings aren't taking effect. I don't see any errors thrown when Claymore starts up except that the cuda library doesn't have a version number.

2) I added a couple of parameters to the Eth command line, but it doesn't seem to pick them up.  For instance, when try to interactively change the dcri it tells me that i need to set -asm 1, but that is already set in the bash file.

Code:

if [ $COIN == "ETH" ]
then
HCD='/home/m1/9.0/ethdcrminer64'
ETHADDR="$ETH_ADDRESS/$ETH_WORKER"

if [ $ETHERMINEdotORG == "YES" ]
then
ETHADDR="$ETH_ADDRESS.$ETH_WORKER"
fi

until $HCD -epool $ETH_POOL -ewal $ETHADDR -epsw x -mode 1 -esm 0 -estale 0 -asm 1 -dcri 12
   do
   echo "FAILURE; reinit in 5" >&2
   sleep 5
done
fi

Any recommendations would be much appreciated.  Windows 10 sucks for mining - so much babysitting, so hoping to get this tweaked and running well.

Powerlimits in windows are in percent TDP; in linux they are in watts.  Thus a powerlimit of 70 in windows is .7 * 120 watts or whatever the TDP is (84 watts if the card TDP is 120); while 70 in linux is 70 watts.  I would recommend comparing Windows and linux powerlimits with a kill-a-watt to get them exactly equal; but most likely a 3gb 1060 has a 120 or 125 watt TDP.

Also linux OC offsets are scaled differently than windows; you will need to use higher offsets to get the same results in linux.

In regards to Claymore ETH settings; the Claymore readme says -asm is for AMD only.  I haven't tested this but -dcri 30 (or any valid number) is only supposed to work with mode 1 (ETH only) if -asm mode is used.

[fullzero]

Anyone have any suggestions to this problem, I am all ears.

I have one of my rigs which uses a Asus Prim z270-a when I put the 6th GPU in it will not POST to the bios screen therefore not start up.

From troubleshooting it only happens if I have 2 GPUs plugged into the X16 slots, if I only have one it works fine.

I have changed the BIOS to the recommend from the thread here for the z270-a and set it all to GEN2.

Any other suggestions?

There were 4 Gen2 settings I needed to change, did you change them all? Could be an issue running 2 on the board with 7 or more, i've only tested with 1 on the board +6 risers or all risers. Check for faulty riser cables too, that slowed up a friend of mine getting 7 going on this board.

This was my thought as well; WarwickNZ.

I'm not sure if you can use 2x GPUs direct on the mobo + 5x with risers with this mobo.

When using a GPU directly more pcie lanes / pcie bandwidth is used.  This may result in insufficient bandwidth on one of the cards to enable normal operation.

While using a card directly I would change the pcie bandwidth to gen2 for all cards.  Then when you get your risers, I would change it back to auto. 

As WarwickNZ indicated; there are multiple pcie bandwidth related settings in the bios.  There may also be an SLI setting that needs to be disabled as well.

I should probably rephrase. I have all 6 GPUs on risers, the only time it does not post is when I have 2 of the x16 slots with the risers in them. I will double check and make sure they are all the same gen type. I had it on auto first but when I did, it wouldn't even boot past one card. I'm to the point I may just reset the MB bios via cmos and start fresh.

not sure if you have tried this:

Ensure the monitor is connected to the primary GPU ( the one in the 16x slot closest to the CPU )

Disconnect the USB or SSD/HHD from the rig. 

Fully power off everything: including the PSU.

Press the power button several times to clear any remaining power in the mobo.

Turn the PSU powerswitch back to | "on".

power on (without the USB attached)

See if the bios posts; if you get nothing in 20 seconds; press ctrl + alt + del repeatedly until the system reboots. 

Wait and see if the bios posts.

If the bios posts attach the USB key and press ctrl + alt + delete.