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.
6152 Discussions

Is there a Tool to measure load on Myriad devices

Siegfried
Beginner
503 Views

Load of an Movidius Extension.

Can we measure the load of a Movidius Extension? For example like the CPU load with Process Explorer.

0 Kudos
1 Solution
Peh_Intel
Moderator
331 Views

Hi Siegfried,

 

Changing the settings of the Service Configuration File might fulfil your needs.

 

Modify the hddl_service.config file in <OV>\3rdparty\hddl\config\hddl_service.config and update change "device_snapshot_mode": "none", to "device_snapshot_mode": "full".

 

After changing a configuration file, restart the hddldaemon in <OV>\3rdparty\hddl\bin and run your application with -d HDDL from another window. 

 

The following YouTube video walks you through the process, however, it's for an older release and some parts may have changed.

https://www.youtube.com/watch?v=nyBvFF_z8Sw

 

 

Regards,

Peh

 

View solution in original post

5 Replies
Peh_Intel
Moderator
439 Views

Hi Siegfried,

 

Thanks for reaching out to us.

 

If the “load” you mean is the memory consumption of loading network, then the memory consumption can be measured for HDDL plugin through configuration parameter (HDDL_DEVICE_MEMORY_USED). However, this configuration parameter is not available for MYRIAD plugin (Intel® Neural Compute Stick 2).

 

HDDL_Memory.JPG

 

 

Regards,

Peh

 

Siegfried
Beginner
428 Views

Hi,

Thank you for the answer. Good to know how to see memory usage. But i hoped there is something for computation load. Lets say there are lots of asynchrounous requests runnung (or scheduled) on the VPU. Is one parallel reuqest enough to occupy the whole unit, or can i run 10 requests in parallel? How can we measure what the limit are?

We have a system which starts inferences on demand. Sometimes this could be a high number of inferences, somtimes its just a few requests. We wanted to have a way to debug this and see how hard the vpu is working right now. The only thing we see is, how long it takes until the inference is finished.

Peh_Intel
Moderator
332 Views

Hi Siegfried,

 

Changing the settings of the Service Configuration File might fulfil your needs.

 

Modify the hddl_service.config file in <OV>\3rdparty\hddl\config\hddl_service.config and update change "device_snapshot_mode": "none", to "device_snapshot_mode": "full".

 

After changing a configuration file, restart the hddldaemon in <OV>\3rdparty\hddl\bin and run your application with -d HDDL from another window. 

 

The following YouTube video walks you through the process, however, it's for an older release and some parts may have changed.

https://www.youtube.com/watch?v=nyBvFF_z8Sw

 

 

Regards,

Peh

 

Siegfried
Beginner
303 Views

Hi Peh,

 

Thank you. util% looks very promising. 

 

Best regards,
Siegfried

Peh_Intel
Moderator
282 Views

Hi Siegfried,


This thread will no longer be monitored since we had answered your question. If you need any additional information from Intel, please submit a new question. 



Regards,

Peh


Reply