AMD Ryzen AI Software is a developer-focused toolkit for running machine learning models locally on laptops that support Ryzen AI. It helps you take pretrained PyTorch or TensorFlow models and adapt them to AMD hardware acceleration.
Optimize inference for Ryzen AI
The package includes runtime libraries and utilities for inference optimization. You can configure compute graphs, select target devices, and gain performance from the dedicated AI block. This reduces load on the CPU and GPU and improves responsiveness for on-device models.
Fit into existing ML workflows
Ryzen AI Software is designed to work with established PyTorch and TensorFlow pipelines, supporting model porting without a full rewrite. This makes it easier to prototype and run local assistants, generative models, and computer vision systems directly on a laptop.
Built for AI PCs
The toolkit targets the “AI PC” class of devices, where inference runs on a dedicated neural accelerator. It provides documentation, examples, and libraries to help you bring up and debug models on compatible Ryzen systems.
