Hardware Requirements Matrix
Overview
This document provides the minimum and recommended hardware requirements, along with other dependencies, for running RLC Pro AI effectively. RLC Pro AI is available for the x86_64 and aarch64 architectures.
Requirements Matrix
| Component | Minimum | Recommended |
|---|---|---|
| Memory | 4 GB RAM | 16 GB RAM or more for AI workloads |
| Storage | 20 GB available disk space | 50 GB or more (model files grow quickly) |
| CPU | 4 cores | 4 cores or more |
A base RLC Pro AI cloud image installation occupies roughly 13 GB of disk, so plan storage headroom for models, datasets, and container images on top of that.
GPU Support
A GPU is optional. RLC Pro AI runs AI workloads on the CPU out of the box, and the pre-installed PyTorch build automatically uses a supported NVIDIA data center GPU when one is present.
- NVIDIA: The NVIDIA data center driver (open kernel modules), CUDA Toolkit, cuDNN, NCCL, Fabric Manager, and IMEX ship pre-integrated with the operating system.
- CPU-only: Fully supported. No configuration changes are required to run without a GPU.
Other Dependencies
- UEFI or BIOS-compatible system. Secure Boot is supported through a CIQ-signed shim.
- Network connectivity to
depot.ciq.com(HTTPS, port 443) for package updates and entitlements. - A CIQ Portal account and access token for Depot enrollment. See the Post-Installation Guide.