Conky nVidia temperature reading with Wayland?

I’ve been tinkering with Wayland & nVidia (yes I know I like punishment…) I’ve got it working very well, except for GWE not working & no output for nVidia sensors in Conky. I’m not troubled without GWE (don’t O/C my card anyway), but I do like to keep track of temps because I do BOINC work.

In doing research—I see that nv-control functionality is being added to the NVML Library ( see: https://github.com/NVIDIA/open-gpu-kernel-modules/discussions/262 ). Has anyone here looked into this & in particular looked at the API info to see if sensor reading can be currently enabled? The API information: https://docs.nvidia.com/deploy/nvml-api/index.html

I’ve been looking at it (some heavy lifting involved for me…) & still am not seeing a good way to integrate what it touches on into Conky…

I can look at nVidia settings & it returns a core temp reading—so there must be a variable that we can access…

Thoughts anyone?

nvidia-smi -q -d temperature

==============NVSMI LOG==============

Timestamp                                 : Sun Jan 15 20:22:04 2023
Driver Version                            : 525.78.01
CUDA Version                              : 12.0

Attached GPUs                             : 1
GPU 00000000:01:00.0
    Temperature
        GPU Current Temp                  : 35 C
        GPU Shutdown Temp                 : 101 C
        GPU Slowdown Temp                 : 98 C
        GPU Max Operating Temp            : 87 C
        GPU Target Temperature            : N/A
        Memory Current Temp               : N/A
        Memory Max Operating Temp         : N/A

Should do it.

nvidia-smi -q -d temperature | grep -i "GPU Current" 

Why?

Habit. It is an expansion:

abbrev-alias -g gg="| grep -i "

But I copied and pasted, so case insensitivity not needed :smiley:

Thanks!! I had not thought about nvidia-smi…

${execi 10 nvidia-smi -q -d temperature | grep -i "GPU Current" }
${execi 10 nvidia-smi -q -d temperature | grep -i "GPU Slowdown" }
${execi 10 nvidia-smi -q -d temperature | grep -i "GPU Shutdown" }

At least filled the hole that was there…now for nVidia to “really” make the library useable would be to add fan control & frequency readings…

I don’t have nVidia stuff - but I have seen conky code for displaying fan etc for nVidia. Now if only I could remember where I saw it! Apparently it is extractable - should I go looking for it? It could even be in hwmon somewhere…

No problem…the code above works for now. The problem is with nVidia…they are only providing a small sub-set of what is available in xorg.

It was easy to write:

${nvidia temp 0} 
${nvidia memfreq 0}
${nvidia gpufreq 0} 

to get the info I really needed…You see above what it takes now to just get Temps…there is NO access to GPU Freq or Memory freq or temp in Wayland at this time.

I wasn’t sure about Wayland (haven’t found a reason to use it yet), so I’m not too surprised! A VERY knowledgable conky friend of mine has nVidia, so I was aware he got lots more info into conky, but he’s on X still too…

Glad you got at least the minimums going!

And I now see another problem that is a deal-breaker for me with Wayland. I run BOINC 24/7 & when I woke the system up this morning—BOINC was “frozen”. Need to do some checking into why that happened–of course, that is not happening in Xorg…going to bugreport to BOINC about it.

This topic was automatically closed 2 days after the last reply. New replies are no longer allowed.