Monitoring temperature

[Xinef]

Hi,

I'm trying to generate graphs of GPU and CPU temperatures while running benchmarks, and I've run into a small issue - when I prefix commands with:
MONITOR=all
it does register all of the available sensors, however that's a lot more info than I need, and would clutter the results when I make them public on OpenBenchmarking.org. I'd like to just choose a few sensors for monitoring.

Now I'm trying to figure out what I need to set the MONITOR variable to, in order to monitor just temperatures and CPU/GPU/RAM usage.

Here's my output for:
$ phoronix-test-suite system-sensors

Phoronix Test Suite v7.2.0
Supported Sensors For This System

CPU Frequency (CPU0): 3000.00 Megahertz
CPU Frequency (CPU1): 3000.00 Megahertz
CPU Frequency (CPU2): 3000.00 Megahertz
CPU Frequency (CPU3): 3000.00 Megahertz
CPU Usage (CPU0): 40.82 Percent
CPU Usage (CPU1): 84.00 Percent
CPU Usage (CPU2): 60.00 Percent
CPU Usage (CPU3): 100.00 Percent
CPU Usage (Summary): 73.76 Percent
GPU Fan Speed: 20 Percent
GPU Frequency: 1242 Megahertz
GPU Temperature: 37.00 Celsius
GPU Usage: 1 Percent
Drive Read Speed (sda): 0.00 MB/s
Drive Read Speed (sdb): 0.00 MB/s
Drive Write Speed (sda): 0.00 MB/s
Drive Write Speed (sdb): 0.00 MB/s
Memory Usage: 1938 Megabytes
Swap Usage: 7 Megabytes
System Iowait: 0.00 Percent
System Temperature: 34.38 Celsius

Unsupported Sensors For This System

- Ambient Temperature
- CPU Fan Speed
- CPU Power Consumption
- CPU Temperature
- CPU Voltage
- GPU Power Consumption
- GPU Voltage
- Drive Temperature
- Network Usage
- System Fan Speed
- System Power Consumption
- System Voltage



So far I've figured out (through trial and error) that the following gives me most of what I need:
MONITOR=gpu.temp,cpu.usage.summary,gpu.usage,memor y.usage

However, I still can't figure out the label for System Temperature. I tried:
system.temperature, system.temperature.monitor, system.temp, system.t, system.temperature0, system.temp0, and probably a few other combinations, to no avail...
(when running MONITOR=all, it is labeled System Temperature Monitor on the graph).

I couldn't find anything specific in the documentation, on the web, nor in the forums, so help would be appreciated

______
PS.: I'm running those tests because I'm trying to build a silent PC, which can run under full load for hours, while maintaining reasonable temperatures, yet have fans spinning slow enough to remain inaudible. Thus, I'm using only 140 mm fans, 9 total (6 case, 2 CPU cooler, 1 PSU fan), and a passive GTX 1050 Ti KalmX graphics card, and trying out which fan configuration / speed profile will result in best cooling while maintaining the lowest noise.

[Michael]

You can do MONITOR=all.usage or here are some other examples: MONITOR=cpu.usage,gpu.temp,cpu.temp,gpu.usage etc

[Xinef]

... eh... now moments after posting this, I realized there's one more thing I haven't tried... and it works:

sys.temp

I guess whoever wrote the code likes three letter acronyms

[Michael]

... eh... now moments after posting this, I realized there's one more thing I haven't tried... and it works:

sys.temp

I guess whoever wrote the code likes three letter acronyms

Nope, don't remember the reason I used three letters for the MONITOR var... Though I could rather easily make it more liberal in what it allows for input to match, so say system.temperature.