Add FP64 precision support in intel_npu

[CailiWuIntel]

Details:

Add FP64 precision pipeline support in intel_npu。

Tickets:

• E-121878

[CailiWuIntel]

RUN_CI_ALL

[ArtemySkrebkov]

There are no triggers like RUN_CI_ALL for OpenVINO repo.

Also can you please update the description to avoid using a direct link? We normally post just number like this E-121878

[CailiWuIntel]

There are no triggers like RUN_CI_ALL for OpenVINO repo.

Also can you please update the description to avoid using a direct link? We normally post just number like this E-121878

Thanks for your kindly reminder, the related ticket description is updated.

[PatrikStepan]

Can you please enable the relevant functional tests (for infer_request) as well in vpux-plugin repository?

[CailiWuIntel]

Can you please enable the relevant functional tests (for infer_request) as well in vpux-plugin repository?

That's OK, I will add the test case for fp64 in this PR, but it should be added after this PR merged, because this test pass need fp64 supporting in OV, and I cannot merge my PR8666 without test pass. I've tried to add the test case locally, and it passes, detail shows in comment https://github.com/intel-innersource/applications.ai.vpu-accelerators.vpux-plugin/pull/8666#issuecomment-2071605773

[CailiWuIntel]

Hi @ArtemySkrebkov @StefaniaHergane, could you help to start CI, or tell how can I start it? Also, could you help to merge this PR after CI pass, or tell me what else should I do?

[ArtemySkrebkov]

Started the CI checks.

[iefode]

Have started OV Jenkins CI pipeline

[PatrikStepan]

The relevant NPU functional tests are now part of OpenVINO repository as well: https://github.com/openvinotoolkit/openvino/tree/master/src/plugins/intel_npu/tests/functional/shared_tests_instances/behavior/ov_infer_request Can we enable BehaviorTests/OVInferRequestCheckTensorPrecision*type=f64 now?

[CailiWuIntel]

The relevant NPU functional tests are now part of OpenVINO repository as well: https://github.com/openvinotoolkit/openvino/tree/master/src/plugins/intel_npu/tests/functional/shared_tests_instances/behavior/ov_infer_request Can we enable BehaviorTests/OVInferRequestCheckTensorPrecision*type=f64 now?

I've checked https://github.com/openvinotoolkit/openvino/tree/master/src/plugins/intel_npu/tests/functional/shared_tests_instances/behavior/ov_infer_request/io_tensor.cpp and https://github.com/openvinotoolkit/openvino/tree/master/src/plugins/intel_npu/tests/functional/shared_tests_instances/skip_tests_config.cpp, now the f64 precison test has been added and do not be skipped

So that, it's no need to add any test in this PR.

[CailiWuIntel]

Hi, @iefode @PatrikStepan, could you help to check and start the CI.

[CailiWuIntel]

Hi @ArtemySkrebkov, could you help to start CI.

[ArtemySkrebkov]

Sure, started the CI checks.

[ArtemySkrebkov]

build_jenkins