Base containers for various heterogeneous accelerators and their respective runtimes
This project aims to provide a range of multi-arch base containers that support pre-configured runtimes for various heterogeneous accelerators.
Containers are published on Docker Hub under the acceleratorbase namespace:
| Image | Path | Architecture(s) | Description |
|---|---|---|---|
acceleratorbase/ncs2-rpi |
intel-ncs2/ncs2-rpi | arm64 | OpenVINO runtime for the Intel Neural Compute Stick 2 running on a Raspberry Pi 3 or 4 (raspbian) |
acceleratorbase/openvino-base |
intel-ncs2/openvino-base | amd64 | OpenVINO runtime for the Intel Neural Compute Stick 2 running on AMD64 (ubuntu 20.04) |
acceleratorbase/edgetpu-std |
edgetpu/edgetpu-std | amd64, arm64 | EdgeTPU runtime running at standard clock speed |
acceleratorbase/edgetpu-max |
edgetpu/edgetpu-max | amd64, arm64 | EdgeTPU runtime running at maximum clock speed |
Some examples using the base containers are available in the examples directory.
Please file feature requests (including support for additional target architectures) and bugs in the issue tracker.
This project has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No 825480 (SODALITE).
Licensed under the terms of the Apache 2.0 license, the full version of which can be found in the LICENSE file included in the distribution.