Intel® Distribution of OpenVINO™ Toolkit
Community assistance about the Intel® Distribution of OpenVINO™ toolkit, OpenCV, and all aspects of computer vision-related on Intel® platforms.
6404 Discussions

Receiving MVNC_NO_DATA on mvncGetDeviceOption (MVNC_THERMAL_STATS)

idata
Employee
613 Views

Hi all!

 

Please, can anyone help with the following problem.

 

I always receive the same error MVNC_NO_DATA when I try to call mvncGetDeviceOption with option = MVNC_THERMAL_STATS.

 

Other options seems to work properly. I can get MVNC_TEMP_LIM_LOWER = 85.

 

I tried this on Ubuntu (Virtual Box under Windows) with ncsdk v1.12.1.1.

 

See screenshots.

 

Does anyone know how to get Movidius current temperature, using ncsdk v1?
0 Kudos
4 Replies
idata
Employee
387 Views

Hi @MaxShtaba

 

You will need to do an inference to the device to be able to read the THERMAL_STATS. Insert the following lines the example, <NCSDK INSTALL DIR>/examples/caffe/GoogLeNet/run.py

 

#Insert in line 86 thermal_stats = device.GetDeviceOption(mvnc.DeviceOption.THERMAL_STATS) print("Device Temperature: ", thermal_stats)

 

Once you modify the example code, run the example with make run

 

Hope this helps!

 

Regards,

 

Jesus
0 Kudos
idata
Employee
387 Views

Thank you, @Jesus_at_Intel.

 

That example works fine with this lines.

 

Now I am trying to get this value on Windows using APIv1 or any other way, as I mentioned in another ticket:

 

https://ncsforum.movidius.com/discussion/1584/current-movidius-temperature-on-windows#latest
0 Kudos
idata
Employee
387 Views

As I understand, this GoogLeNet example uses python version of APIv1 library, right?

 

Does it mean, that C/C++ version of APIv1 does not work with the THERMAL_STATS parameter?

 

Or it depends on other calls, maybe call of LoadTensor function?

 

So, If I will try to use C/C++ version of APIv1 with LoadTensor before I try to read temperature, I can get correct result instead of MVNC_NO_DATA?
0 Kudos
idata
Employee
387 Views

Well, there is no more MVNC_NO_DATA error, it really depends on LoadTensor function call.

 

Other problem I will describe in another ticket.

 

Thank you.
0 Kudos
Reply