Datasheet Kubernetes.pdf

Managed Kubernetes for Superclusters

Lambda’s fully-managed Kubernetes offering builds on the power of our Superclusters. We install and maintain a Kubernetes installation on your cluster, run most daily operational tasks, and handle cluster and hardware failures - allowing you and your team to focus on using the cluster rather than operating it.

Operational Ease /

Lambda’s world-class engineering team will deploy, run and operate our reference Kubernetes architecture designed for AI, machine learning and GPU workloads. This includes:

In return for allowing us to take full control of the hardware, you will get a fully-managed experience, with access to the Kubernetes API, graphical dashboard, and other components pre-configured:

Single-Tenant Security /

Lambda Superclusters are single-tenant and all Kubernetes and fleet management components will remain local. Nothing inside of the cluster will be exposed to the Internet unless you choose to - by default, all access is over a secure VPN, including to the Kubernetes API and dashboard. Lambda maintains access in order to run monitoring, alerting, and perform proactive fixes when we notice issues.

Single-tenancy also means you get flexibility as to when we do upgrades and to which versions; we’ll work with you to establish maintenance windows and your preferred upgrade cadence.

We perform any required critical upgrades to cluster components, and we proactively monitor your cluster 24/7/365 and attempt to fix issues as soon as possible - resolution times vary by issue type and severity. All clusters are located in datacenters where Lambda have 8x5 continuous presence and 24x7 on-call availability.