The News: Google launched GKE Enterprise, a premium version of its Google Kubernetes Engine managed service, at Google Cloud Next ’23. You can read more details of the announcement on the Google Cloud blog.

Google Enhances GKE With Advanced Security, “Cluster Fleet” Management

Analyst Take: Kubernetes, an open-source container orchestration platform, has its roots deeply embedded in Google’s internal infrastructure. The origin story of Kubernetes dates to Google’s internal project called “Borg,” a large-scale cluster manager developed to handle Google’s global-scale billion user plus services such as Gmail and YouTube. Borg’s architectural principles, along with its operational experience, laid the foundation for Kubernetes.

In mid-2014, Google decided to bring these previously internally developed advancements to the broader community by open-sourcing Kubernetes. The initial release was developed by Joe Beda, Brendan Burns, and Craig McLuckie, who took key ideas from Borg, but Google Kubernetes Engine (GKE) is Google’s commercial version of Kubernetes, and forms the bedrock of Google container offerings. GKE Enterprise combines GKE, a service for running containers solely in the Google Cloud, and Google Anthos for running Kubernetes in hybrid and multi-clouds. The Enterprise version improves Kubernetes management through additional security, governance, service mesh management, a dashboard to see all workloads, and “cluster fleets.”

GKE Enterprise edition incorporates Google’s Anthos multi-cluster fleet management capabilities. That approach lets platform engineers group similar workloads into dedicated clusters, apply custom configurations and policy guardrails per fleet, isolate sensitive workloads, and delegate cluster management to other teams. GKE Enterprise’s managed security features include advanced workload vulnerability insights, governance and policy controls, and managed service mesh. Hybrid and multi-cloud support in GKE Enterprise allows customers to run container workloads anywhere GKE, in other public clouds, or on-premises with Google Distributed Cloud.

GKE Enterprise edition will be available in preview in early September, with launch partners Accenture, CDW, Deloitte, DoiT International, SADA, Searce, and 66degrees.

GKE Enterprise also tied into other Google Next launches. Users can run Cloud TPU workloads on GKE to take advantage of features such as autoscaling, workload orchestration, and support for 15,000 node clusters when developing AI models. Google enhanced Cloud TPU at the conference with version v5e, with up to 2x higher training performance and 2.5x higher inference performance per dollar for large language models (LLMs) and gen AI models compared to Cloud TPU v4.

When running workloads in GKE, customers can save commute cycles by scaling up and down according to demand because they only pay for TPU resources provisioned.

Other GKE announcements at Google Next included:

GKE now supports A3 VM with NVIDIA H100 GPU for training large models
Cloud Storage FUSE is in general availability on GKE for moving unstructured data workloads to GKE without changing how that data is accessed
The ability to run Duet AI and Cloud Run in GKE is available in preview

The Futurum Group’s research shows an overwhelming majority (84%) of customers want a container management system that is integrated, maintained, and supported by a commercial vendor. Also, the biggest challenges to Kubernetes adoption are lack of talent to manage the new technology and difficulty keeping people who do have that expertise.

Those trends work in Google’s favor because it offers container management as a commercial service integrated with other services. We see GKE Enterprise as an extension of Google’s rich Kubernetes and AI/ML capabilities. Google is already the industry leader in Kubernetes, and combines that value with a robust suite of AI/ML and data services to support rapid innovation. In this case, it has already impressively integrated GKE Enterprise with its other new AI services at launch.

Looking Ahead

The market for commercial container orchestration and management solutions is led by Red Hat with its OpenShift solution, with the likes of SUSE with Rancher Prime also figuring into the market equation, especially in Europe. As Kubernetes becomes the default for modern hybrid multi-cloud management and deployment, the competitive landscape for which vendor leads this vital control plane will be interesting to watch as public cloud deployments eventually outweigh on-premises deployments.

Customers are increasingly facing a choice with how they look to deploy Kubernetes. On one hand, public cloud providers are offering fully managed “enterprise-grade” offerings such as this newly-announced service from Google and the likes of EKS and ECS from AWS, whereas others partner with Red Hat or SUSE and take a platform-neutral approach. We believe that customers should carefully weigh the simplicity and one-stop shopping benefits of public cloud-provided K8S services against the neutrality of third-party offerings as they make what could be the most pivotal deployment choice in their hybrid cloud strategy.

Disclosure: The Futurum Group is a research and advisory firm that engages or has engaged in research, analysis, and advisory services with many technology companies, including those mentioned in this article. The author does not hold any equity positions with any company mentioned in this article.

Analysis and opinions expressed herein are specific to the analyst individually and data and other information that might have been provided for validation, not those of The Futurum Group as a whole.