NVIDIA Run:ai Is Now Open Source as Part of DSX AI Factory
NVIDIA says the goal is to let the whole ecosystem build faster, handing production-grade AI workload and GPU orchestration to customers, developers and partners
Tool Overview
Details, capabilities and usage guidance
Features
- ✓ The full platform, open sourced: NVIDIA says it is making the complete NVIDIA Run:ai platform available as open source software, giving customers, developers and ecosystem partners access to production-grade AI workload and GPU orchestration capabilities
- ✓ Four components adoptable independently: NVIDIA lists KAI Scheduler for GPU-aware Kubernetes scheduling, the GPU Fractioning Operator for CUDA-native GPU sharing, KAI Resource Management for organisational quota and hierarchy, and Workload Map for a consistent workload-management contract
- ✓ Partners are integrating it: NVIDIA says eight DSX ecosystem partners are integrating the open source platform into their offerings - Crusoe, Mirantis, Nebius, OpenNebula Systems, Rafay, Saturn Cloud, Spectro Cloud and vCluster
- ✓ Adopt the whole or just parts: NVIDIA says platform teams and partners can adopt Run:ai as an integrated whole, build on it for their own products and services, or integrate individual components based on their requirements
- ✓ Companion open source projects: NVIDIA says Model Streamer and Grove complement the platform by accelerating model loading and orchestrating multicomponent inference workloads
How to Use
- Read NVIDIA's official technical blog to confirm what the release covers and which four components can be adopted independently
- Follow NVIDIA's pointers to the NVIDIA DSX AI Factory organisation on GitHub to inspect the full source, plus component repositories such as KAI Scheduler and the GPU Fractioning Operator
- To self-host the integrated stack, note that NVIDIA says the complete open source platform is also published on NVIDIA NGC as a Community Product with a Helm chart
- Note NVIDIA's licensing and support terms: the Community Product requires no paid Run:ai subscription or NVIDIA support entitlement, enterprise support is not included, and the source code is governed by its applicable open source licenses
FAQ
What does this open source release include?
NVIDIA says in its official technical blog that it is making the complete NVIDIA Run:ai platform available as open source software, and it lists four components that can also be adopted independently: KAI Scheduler, GPU Fractioning Operator, KAI Resource Management and Workload Map. Together with additional platform code, these form the integrated NVIDIA Run:ai platform.
Why is NVIDIA open sourcing Run:ai?
NVIDIA says the release expands the NVIDIA DSX AI Factory Platform by establishing Run:ai as an open foundation that lets the broader ecosystem bring differentiated solutions to market faster in response to evolving customer needs. NVIDIA also says that, with the platform open sourced, every DSX ecosystem partner and AI cloud can increase workload throughput, improve GPU utilisation and operate shared AI infrastructure more efficiently.
Which ecosystem partners are involved?
NVIDIA lists eight DSX ecosystem partners that are integrating the open source platform into their offerings: Crusoe, Mirantis, Nebius, OpenNebula Systems, Rafay, Saturn Cloud, Spectro Cloud and vCluster.
Can I adopt just one component?
NVIDIA says yes. KAI Scheduler, GPU Fractioning Operator, KAI Resource Management and Workload Map can all be adopted independently, and platform teams and partners can either take Run:ai as an integrated whole or integrate individual components based on their requirements.
How do I deploy it for self-hosting?
NVIDIA says the complete open source NVIDIA Run:ai platform is available on NVIDIA NGC as a Community Product, with a Helm chart for installing the integrated stack.
Does the open source version include enterprise support?
NVIDIA states clearly that downloading the Community Product from NGC requires accepting the applicable NVIDIA Software License Agreement and Product-Specific Terms for NVIDIA AI Products. No paid NVIDIA Run:ai subscription or NVIDIA support entitlement is required for the Community Product, and NVIDIA enterprise support is not included. The source code is governed by its applicable open source licenses.