how to monitor realtime npu power consumption while inferencing with llama on Intel® Core™ Ultra 7.

[Arya_Pulkit]

how can i monitor realtime npu power consumption during inferencing with llama on Intel® Core™ Ultra 7 ?

[JohnM_Intel]

Hello Arya_Pulkit,

Thank you for raising this concern on Intel Community Forum.

You may find the NPU performance via Task Manager, but for us to better understand your concern kindly follow the steps below to provide a screenshot of the performance tab.

1. Open Task Manager in the system been used to reproduce the issue

2. Go to the Performance tab and make sure both GPU and NPU are present.

3. Reproduce the issue and monitor to usage of GPU and NPU in Task Manager (for GPU, please make sure you are monitoring the Compute component and not the 3D component, as marked with the cyan arrow in the image above).

4. Note is the case if the AI feature is sending it's workload to the GPU, the NPU or both.

And also may I know the reason why you need to monitor the NPU? and why do you use llama?

Best Regards,

John Sergio M.

Intel Customer Support Technician

[JohnM_Intel]

Hello Arya_Pulkit,

Good day!

Just want to know if you have the time to answer my question to better understand the issue you want to be resolved, and also I want to ask what the specific Intel Core is Ultra 7 processor is that you have.

Best regards,

John Sergio M.

Intel Customer Support Technician

[JohnM_Intel]

Hello Arya_Pulkit,

Good day!

How are you? Just letting you know that I'm waiting for your response regarding this post for us to further investigate this issue.

Best regards,

John Sergio M.

Intel Customer Support Technician

[JohnM_Intel]

Hello Arya_Pulkit,

Since this thread is inactive, I will now proceed with closing this one.

Thank you and have a great day!

Best regards,

John Sergio M.

Intel Customer Support Technician