OpenVINO Toolkit SDK for Workshop

This SDK provides the Intel OpenVINO™ Runtime, Python bindings, and Intel GPU driver support for AI inference workloads in Workshop. OpenVINO runs on CPU by default and is fully functional without dedicated hardware; the gpu plug enables hardware-accelerated inference on compatible Intel GPUs. arm64 is listed as a build target but setup scripts target amd64; arm64 support is experimental and untested.

---

Reference workshop

A minimal workshop:

# workshop.yaml
name: openvino-app
base: [email protected]
sdks:
  - name: openvino
    channel: 2025/stable

actions:
  verify: |
    python3 -c "from openvino import Core; print(Core().available_devices)"

This demonstrates a minimal OpenVINO environment that verifies the runtime is available.

---

Using the SDK

Prerequisites, project layout

1. No prerequisite SDKs are required.
2. No specific project layout is needed; OpenVINO can be used with any Python or C++ project.
3. On launch, the SDK installs Intel GPU compute drivers from the ppa:kobuk-team/intel-graphics PPA, configures ldconfig for OpenVINO shared libraries, and registers the Python bindings via a .pth file. No user action is needed.

Verify the runtime

Once the workshop is ready:

workshop shell
python3 -c "from openvino import Core; print(Core().available_devices)"

Expected output: ['CPU'] without the gpu plug, or ['CPU', 'GPU'] with it enabled.

To run a self-contained inference example, download a small ONNX model and compile it:

workshop shell
# Download a small test model
wget -q https://github.com/onnx/models/raw/main/validated/vision/classification/squeezenet/model/squeezenet1.0-3.onnx -O model.onnx
python3 -c "
from openvino import Core
core = Core()
model = core.read_model('model.onnx')
compiled = core.compile_model(model, 'CPU')
print('Model compiled successfully on', 'CPU')
"

Check device availability

From within the workshop shell:

workshop shell
python3 -c "
from openvino import Core
core = Core()
print('CPU available:', 'CPU' in core.available_devices)
print('GPU available:', 'GPU' in core.available_devices)
"

---

Plugs (resources this SDK consumes)

gpu

• Interface: gpu
• Purpose: Grants access to Intel GPU hardware on the host for hardware-accelerated inference.

OpenVINO runs on CPU by default and is fully functional without this plug. Adding the gpu plug enables the GPU device for inference. To verify GPU availability:

python3 -c "from openvino import Core; print(Core().available_devices)"

If GPU is detected, GPU appears in the output list alongside CPU.

Slots (resources this SDK provides)

openvino-libs

• Interface: mount
• Workshop source: $SDK
• Purpose: Exposes the entire SDK install tree (OpenVINO libraries, Python bindings, pkgconfig files) to consumer SDKs.

No SDK in the current store consumes this slot. A consumer SDK would declare a matching plug to mount the OpenVINO libraries into its own environment:

# workshop.yaml (illustrative -- no consumer SDK exists today)
name: openvino-consumer
base: [email protected]
sdks:
  - name: openvino
    channel: 2025/stable
  - name: my-inference-app
    channel: all/edge

---

Documentation and guidance

• OpenVINO official documentation
• Workshop documentation

---

Community and support

• OpenVINO community: OpenVINO GitHub
• Intel developer forums: Intel Developer Community
• Workshop forum: Discourse
• Please review our Code of Conduct before participating.

---

Contributions

All contributions, including code, documentation updates, and issue reports, are welcome!

• See CONTRIBUTING.md for guidelines.
• Open issues or pull requests on the official repository.

---

License and copyright

Copyright 2025 Canonical Ltd.

The OpenVINO Toolkit is licensed under the Apache License 2.0.

Intel, the Intel logo, Intel Atom, Intel Core, Intel Xeon Phi, Iris, OpenVINO, the OpenVINO logo, Pentium, VTune, and Xeon are trademarks of Intel Corporation or its subsidiaries.