With the introduction of v9.1, a container will automatically use the H.264 hardware acceleration available on certain GPUs (dedicated or integrated) - this is used when the container decodes the H.264 video coming from your cameras (so that it can compress it for EdgeVis transmission). Older versions will use a software decoder, which uses more CPU.

Performance gains vary depending on your intended output EdgeVis video stream settings.
​
This is because the higher you set your EdgeVis video settings, the higher the proportion of time your container will spend on EdgeVis compression. For lower settings (e.g. 480p), your container CPU usage may see up to a 50% reduction, while at 1080p you may only see a 10-20% reduction.

EVVR Containers will use the Vulkan programming API to access any available H.264 hardware decoders on your system.

This release officially supports the following:

It is highly recommended to use up-to-date graphics drivers and ensure your system is up to date. By default, both of the above work on an out-of-the-box Ubuntu 24 installation.

As v9.1 introduces this feature, if you installed v9.0 and used a Docker Compose file to install the container, you will need to make changes to enable this feature.

You will need to add the new lines that pass through your hardware accelerator from your host to the container.
​

More info: Upgrading your container from v9.0 to v9.1
​

Depending on your use case, you may want to disable hardware acceleration. A common reason is that you are using other third-party containers that would benefit more from the use of the GPU hardware.

There are two ways to disable hardware acceleration:

You have multiple options for configuring your container using a config file - including burning in these settings using environment variables, or using a seed config file to provide initial settings for your users.

To enable/disable hardware acceleration, add the following setting to your config file:

​