LABCOM-2009 - Simplify AI Deployments with Cisco AI Pods
Proctors |
Sergei Pakhomov None |
Vivek Dalvi None |
This lab highlights the capabilities of the Cisco AI PODS for Inferencing solution deployed on the Cisco UCS X-Series Modular System. It is designed to provide users with hands-on experience in using and managing AI workloads in a modular infrastructure.
Participants can explore pre-configured components to understand the system’s setup and functionality. Additionally, users are empowered to install, configure, and modify their own AI workloads, enabling them to tailor the environment to their specific use cases.
This demo provides access to two separate single node Red Hat OpenShift clusters:
1. Shared Cluster: hosting a Mistral-Small-24B-Instruct LLM. with read-only access, enabling users to explore the platform without making changes.
2. Lab Cluster: each user is assigned admin privileges within their own OpenShift AI Project/ Namespace. Users are also allocated a predefined amount of resources, including RAM, CPU, and GPU slices, to support their activities. This cluster also hosts users own instances of Open WebUI and a MistralDB.