GPU Temperature Monitor => for Linux

Hi folks,

Someone know if there is good Monitor, to monitor GPU monitors and shutdown system if they're getting too hot? (for Linux)

If I cannot find this I will write a bash script with lm-sensors... (public).


I just use this to see temps live:

$ watch -n3 sensors

It will output the readings of sensors every 3 seconds to the terminal.

By using tmux I can set up a terminal with several subwindows.
So I have 1 window with the miner running, another one with se sensors readings and usially a third one for the bash to work on config files.

Have not seen a GUI that shows the readings for a rig with several GPUs.


If you want to try ethOS, we have temperature monitoring. Example panel:

It is in active development:

1 Like

That looks cool.

What about hashrates on ethOS? Did not try the OS yet, although I bought it for 39 bucks. :grin:

You can run the latest silentarmy build on ethOS, or install optiminer. We do not use untrusted closed-source software out-of-the-box.

Like Optiminer? :joy:

Just kidding.

SA5 does not perform good enough atm to compete with Claymore. And Optiminer does not work for me as it drops out some of my GPUs after some time. If SA will catch up to Claymore, I'll give ethOS a try! :cookie:

Sounds good. We have a process for inserting the any new/faster SA build, so we are ready...

please back to the topic, I don't want to buy a free OS.

1 Like

You can still get it on torrent. Free software is free. But you have to check the checksums of the downloaded iso to be sure it is not manipulated. Also there could appear a small goat with red eyes in your bedroom, that's part of the license of ethOS. :joy:

@AnorakSupra I don't want to use such software also not for free. This is against my understanding of GNU and free Software.

1 Like

So coded a simple script which shutdown the computer if one card gets higher temp > 90°C.

Comments are welcome ....

# coded 16.11.16 by xfcash
while true; do

  # getting GPU temp as integer, and compare EACH shutdown if temp x > 90 
  for TEMP in $(sensors | awk -v RS= -v OFS=' ' '{for(i=1;i<NF;i+=2) $i=$i":"}1' | grep amdgpu | sed 's/^.*\+\(..\)..°C.*crit.*$/\1/g');
  [ "$TEMP" -gt "90" ] && shutdown -h now
  sleep 60

Cool, thanks for the script.
Have been writing on a script too, but as I use Claymore again I run my rigs on Windows again.
Propably just switching between Linux and Windows (dual boot) for the fastest miner each time.

no problem. Im currently using optiminer v0.3.3 for 1 day now. I've heard cm miner need 20W more per GPU and is currently not much faster.... getting now stable 130H/s each card.

Nice, thx for sharing you scripts!! I still waiting for the arrived of 2 GPUs, but I am undecided about use Win or Linux!

Use psensor

He need a sensors monitoring that can make some task when have some alarms

Uhmm, look this, I think that you can create a custom scripts with this program:


1 Like