Home > AI Solutions > Artificial Intelligence > White Papers > Simplifying AI: Dell Enterprise Hub Enables Kubernetes Deployment for AI Models > Solution approach
Dell Technologies has partnered with Hugging Face to introduce Kubernetes support for all available models on the Dell Enterprise Hub. The introduction of Kubernetes adds another layer of flexibility and scalability on top of a solution that already makes on-premises generative AI model deployment easy. As generative AI models continue to grow and become more complex, Dell Technologies and Hugging Face work ceaselessly to adapt both physical and software infrastructure to the ever-changing world of AI.
Previously, users have been able to utilize Docker support to deploy their generative AI models. However, the lack of Kubernetes support limited the usefulness of Dell Enterprise Hub for many businesses. Dell Technologies and Hugging Face collaborated to release Kubernetes support. Dell Technologies and Hugging Face have tested and validated the models available on Kubernetes to ensure that customers can confidently use Dell Enterprise Hub with a variety of server platforms, GPUs, and AI models.
On top of the support already provided, organizations can utilize the other advantages that Hugging Face provides to subscribers. Features like a single sign-on portal, security features and validation, and on-demand support from Hugging Face experts allow organizations to confidently use AI without the headache. In the context of this paper, Dell Technologies provides a high-level walkthrough of using Dell Enterprise Hub to deploy an inference model on the Kubernetes environment to demonstrate just how easy to use Dell Enterprise Hub is.
The diagram above provides a high-level overview of the Kubernetes architecture Dell Technologies used while validating the Kubernetes deployments. At the bottom, the physical layer exists with networking and compute infrastructure. At the layer above, there exists the operating system and the container runtime with the Kubernetes control plane and Kubernetes workers. The GPU and network operators configured for the Kubernetes workers are essential for utilizing system GPUs and outside networking. Both must be properly configured before an AI Model can be deployed. Finally, the top layer are the models provided by Dell Enterprise Hub on Hugging Face.