After running job on a selected hardware from jupyter, I get to see in the output file with 'JOB_RUNTIME'. is there way to get other stats like cpu usage, memory, and other stats
Thank you for posting on the Intel® communities.
After checking your post, we noticed that this is related to Intel DevCloud and Jupyter Notebook (JupyterLab). We have a forum for those specific issues, questions and products so we are moving this to the Intel® DevCloud Forum so it can get answered more quickly:
Intel Customer Support Technician
Could you please let us know the below details:
- Are you using oneAPI Devcloud or FPGA Devcloud for your work?
- Could you please clarify on how do you get the output file with JOB_RUNTIME parameter as you mentioned?
- If you want to get your application performance in terms of cpu usage and memory, you can make use of Intel Vtune Profiler, one of the best performance analysis tool provided by Intel. You can use Intel® VTune™ Profiler Application Performance Snapshot for a quick view into different performance metrics like cpu usage, memory etc. Please find the below documentation for more information.
Thanks for the response. I am running on cpu on "devcloud for edge" with jupyter.
I see output of script in the end as below :
skipping application metrics
This forum is intended to handle only oneAPI Devcloud issues. Since you are working on Devcloud for edge, we are moving this query to Intel Devcloud for edge forum for a faster response.
Thanks for reaching out.
Yes you can use the telemetry dashboard. Please check the Advanced Sample Applications (e.g. Accelerated object detection) for more information about how to use the telemetry dashboard.
Thanks for the response.
It is not clear or have no mention of use of telemetry dashboard in Advanced Sample Applications (e.g. Accelerated object detection) . If you can please share some info on this or any document on how to use it will be very helpful.