How to achieve memory sharing for NPU with openvino?

[WilsonChen0723]

Hi,

I'm trying to achieve memory sharing for NPU with openvino to get better processing time.

I've done this before with GPU with your help in this post.

My question is : How to achieve the same thing with NPU to implement zero copy?

Will Remote Tensor API of GPU Plugin support NPU?

BR

Wilson Chen.

[Aznie_Intel]

Hi Wilson Chen,

Thanks for reaching out.

Remote Tensor API is only supported with GPU plugin. The NPU support in OpenVINO is still under active development and may offer a limited set of supported OpenVINO features. For NPU device, you may check the Supported Features and properties from the documentation.

Regards,

Aznie

[Aznie_Intel]

Hi Wilson Chen,

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

Regards,

Aznie