SUSE has announced a new wave of capabilities for SUSE Rancher Prime and SUSE Virtualization, aiming to push two specific ideas: utilizing intelligent AI agents to simplify infrastructure operations and bringing management of virtual machines and containers closer together. The company presented these updates on March 24 and frames them as part of an evolution toward more automated operations, with richer context and less reliance on manual tasks in hybrid and cloud-native environments.
The announcement comes at a time when many organizations are simultaneously reassessing their Kubernetes strategies, virtualized infrastructure, and preparations for AI workloads. SUSE seeks to position itself precisely at this intersection: offering a common management layer for containers, VMs, and platform operations, with a clear focus on open tools and more integrated AI in the daily work of SRE and operations teams.
Liz evolves from assistant to an “AI team” within Rancher Prime
The most striking innovation is Liz, SUSE Rancher Prime’s AI assistant. Until now, it functioned as a contextual aid within the platform, but SUSE has decided to expand it into a multi-agent architecture. According to the company, Liz now coordinates specialized agents in observability, security, Linux, provisioning, and fleet management, allowing users to operate through a single interface while the system distributes tasks among more specific agents. SUSE claims this approach reduces cognitive load and speeds up tasks such as deployment, platform health management, and incident investigation.
This change has another important implication: opening up to external services via the Model Context Protocol (MCP). SUSE states that Liz can now directly connect to third-party MCP servers and incorporate those external sources into its “AI Crew.” In practice, this would enable linking Rancher Prime with tools like Atlassian, internal documentation, or proprietary services without bespoke integrations. The company presents this as the first open ecosystem of AI agents for such operations.
It’s worth adding a nuance here. While the commercial messaging is ambitious, SUSE is primarily talking about expanding operational context and automating interactions with the platform—not replacing technical teams. In fact, their materials emphasize that the goal is to augment engineers’ capabilities, not eliminate them. This aligns with the tasks they highlight: troubleshooting, visibility, recommendations, and operational coordination within Rancher and Kubernetes.
More unified virtualization with a clear focus on enterprise AI
The other major aspect of the announcement concerns SUSE Virtualization. The company enhances this line with support for NVIDIA MIG, a GPU partitioning technology that allows a single GPU to be divided into multiple isolated instances. Practically, this improves hardware utilization for AI workloads and enables more efficient multi-tenant scenarios—critical as GPU costs remain high and organizations seek to maximize each equipped node. SUSE confirms that MIG vGPU support has been generally available in SUSE Virtualization 1.7 since January 2026 for NVIDIA A100, H100, and H200 families.
Additionally, the company introduces new operational tools such as VM Auto Balance, Live Storage Migration, and more granular update controls. SUSE differentiates between features already available and those still maturing: NVIDIA MIG support and update controls are available now, while VM Auto Balancing and Live Storage Migration are in Early Access. This detail helps set realistic expectations around the current state of each capability.
Beyond specific functions, the core message is clear: SUSE aims to reinforce the concept of an integrated platform for VMs and containers, especially useful for companies that cannot or do not want to move everything to a single operational model. The company acknowledges that containers are now the standard for many modern and AI workloads but also recognizes that virtual machines remain essential in enterprise environments. This balance is, in fact, a key market aspect SUSE is aiming to leverage.
More self-service for developers and better GPU utilization
SUSE also used this announcement to bolster its developer support. Rancher Prime provides immediate access to part of its catalog of over 140 hardened, enterprise-ready applications, including components like Base Container Images, PostgreSQL, Redis, and Penpot. The goal is to promote a “shift left” approach to supply chain security, encouraging developers to start from more trusted images and components.
Alongside this, the company spotlighted Virtual Clusters with multi-tenancy for GPUs. SUSE describes these as large-scale sandbox environments where teams can provision isolated and self-service Kubernetes control planes to experiment with AI models and complex workloads without affecting the rest of the organization. SUSE’s documentation notes that these environments can detect available GPUs on the host and expose them to developers via a self-service mechanism—reducing administrative bottlenecks. These Virtual Clusters are claimed to be production-ready and generally available.
Overall, this shift reinforces a well-known strategy. SUSE is not just competing with a message focused solely on Kubernetes or virtualization but offering a broader open infrastructure for modern workloads, where AI, containers, VMs, and operations converge. The 2026 update emphasizes that this convergence isn’t just about unified management but also about operational automation supported by contextual AI and optimized use of expensive resources, especially GPUs.
Frequently Asked Questions
What exactly has SUSE announced for Rancher Prime?
SUSE has expanded Rancher Prime with a multi-agent architecture for Liz, its AI assistant, and with MCP integration for external data sources and third-party tools, eliminating the need for custom integrations.
What does NVIDIA MIG bring to SUSE Virtualization?
It allows partitioning enterprise GPUs into multiple isolated instances, improving hardware efficiency for AI workloads and enabling multi-tenant scenarios. SUSE states that this capability is generally available in SUSE Virtualization 1.7.
Which features are already available and which are still in Early Access?
SUSE reports that AI capabilities in Rancher Prime, the AI Assistant improvements, NVIDIA MIG multi-tenancy, Upgrade controls, and Virtual Clusters are available now, while VM Auto Balancing and Live Storage Migration are in Early Access.
Who is this SUSE announcement most relevant for?
Primarily organizations operating both Kubernetes and virtualization, looking to incorporate AI into platform operations and better leverage GPU resources in enterprise and hybrid environments.
via: suse