Supported Integrations
NVIDIA Run:ai includes built-in support for the following Kubernetes components and external tools. You may still need cluster-side installation (for example, install an operator and its CRDs) before you can use the integration.
Frameworks
| Framework | Category | Supported Version | Additional Information |
|---|---|---|---|
| Dynamo operator | Distributed inference | 1.5.x | Dynamo operator is a Kubernetes operator that simplifies the deployment, configuration, and lifecycle management of DynamoGraphs. NVIDIA Run:ai provides out of the box support for submitting Dynamo workloads via YAML. See Dynamo operator documentation for more details. |
| NIM operator | Model Serving | 3.0.x | The NVIDIA NIM Operator enables Kubernetes cluster administrators to operate the software components and services necessary to deploy NVIDIA NIMs and NVIDIA NeMo microservices in Kubernetes. NVIDIA Run:ai provides out of the box support for submitting NIM operator workloads via YAML. See NIM operator documentation for more details. |
| LeaderWorkerSet (LWS) | Distributed inference | 0.6.0 or higher | NVIDIA Run:ai provides out of the box support for submitting LWS workloads via YAML and NVIDIA Run:ai native distributed inference workloads using LWS. |
| Kubeflow MPI | Distributed training | MPI Operator v0.6.0 or higher | NVIDIA Run:ai provides out of the box support for submitting MPI workloads via API, CLI or UI. See Distributed training for more details. |
| PyTorch | Distributed training | Kubeflow Training Operator v1.9.2 | NVIDIA Run:ai provides out of the box support for submitting PyTorch workloads via API, CLI or UI. See Distributed training for more details. |
| TensorFlow | Distributed training | Kubeflow Training Operator v1.9.2 | NVIDIA Run:ai provides out of the box support for submitting TensorFlow workloads via API, CLI or UI. See Distributed training for more details. |
| XGBoost | Distributed training | Kubeflow Training Operator v1.9.2 | NVIDIA Run:ai provides out of the box support for submitting XGBoost via API, CLI or UI. See Distributed training for more details. |
| JAX | Distributed training | Kubeflow Training Operator v1.9.2 | NVIDIA Run:ai provides out of the box support for submitting JAX workloads via API, CLI or UI. See Distributed training for more details. |
| Triton | Orchestration | Any version | Usage via docker base image |
Development Tools
| Tool | Category | Additional Information |
|---|---|---|
| Jupyter Notebook | Development | NVIDIA Run:ai provides integrated support with Jupyter Notebooks. See Jupyter Notebook quick start example. |
| PyCharm | Development | Containers created by NVIDIA Run:ai can be accessed via PyCharm. |
| VScode | Development | Containers created by NVIDIA Run:ai can be accessed via Visual Studio Code. You can automatically launch Visual Studio code web from the NVIDIA Run:ai console. |
Storage and Registries
| Tool | Category | Additional Information |
|---|---|---|
| Docker Registry | Repositories | NVIDIA Run:ai allows using a docker registry as a Credential asset |
| GitHub | Storage | NVIDIA Run:ai communicates with GitHub by defining it as a data source asset |
| S3 | Storage | NVIDIA Run:ai communicates with S3 by defining a data source asset |
Experiment Tracking and Monitoring
| Tool | Category | Additional Information |
|---|---|---|
| TensorBoard | Experiment tracking | NVIDIA Run:ai comes with a preset TensorBoard Environment asset |
Infrastructure and Cost Optimization
| Tool | Category | Additional Information |
|---|---|---|
| Karpenter | Cost Optimization | NVIDIA Run:ai provides out of the box support for Karpenter to save cloud costs. Integration notes with Karpenter can be found here. |