The Rescale AI Physics Local Inference App gives engineers a native desktop application for Windows, Mac, and Linux that works directly with AI models trained on the Rescale platform. Engineers can browse available models, install lightweight inference packages, and run predictions on their own hardware without an active cloud session.

The app runs on CPU by default and automatically takes advantage of local GPU hardware when available, with no requirements for additional libraries or frameworks. This is particularly useful during design review cycles, when a CAE engineer needs to query a surrogate model interactively without queuing a cloud job. It also supports local inference workflows for on-prem teams and facilities with network restrictions.

In a digital engineering context, this is particularly useful during design review cycles, when a designer or CAE engineer needs to query a surrogate model interactively without queuing a cloud job. It also unlocks local inference workflows for on-prem teams or facilities with network restrictions.

Watch a Demo and learn more about the latest AI physics capabilities on Rescale.