Requirements

Software

OS: Microsoft Windows 10+, Linux CentOS 7+, Linux Rocky 8+, Linux RHEL 8+ (64-bit).
CPU: 64-bit Intel or AMD multi-core processor.
GPU: NVIDIA GPU suggested for accelerated ML workflows, with 6GB of VRAM or more recommended. CPU fallback is supported when GPU execution is not available, but it can be slower.
RAM: 8GB required (16GB or more recommended).
Storage: At least 700MB of free disk space for the standard installation. If ML dependencies are required, then at least 7GB of free disk space.

NOTE

Adonis may work with other CPU, GPU, RAM, and VRAM specifications. The list above represents the specifications that have been tested successfully.

NOTE

Internet connection is required for licensing and ML dependency download only.

With an AdonisFX license, see Licensing for details, installing the ML dependencies is optional. It is only required to run ML inference on the GPU. If not installed, ML inference runs on the CPU.

AdonisML features require an AdonisML license bundle. See Licensing for details.
ML features run locally and do not require cloud services. Training data remains on the local machine.
Generating ML training data and training models may require additional disk space and memory depending on the dataset and model size.
Trained models are generally expected to be under 2GB, but extracted training data may exceed 10GB depending on the scenario.