{"id":327049,"date":"2026-04-30T22:00:02","date_gmt":"2026-04-30T16:30:02","guid":{"rendered":"https:\/\/ebiztoday.news\/?p=327049"},"modified":"2026-04-30T22:00:02","modified_gmt":"2026-04-30T16:30:02","slug":"kubernetes-control-plane-and-red-hat-drive-ai-infrastructure","status":"publish","type":"post","link":"https:\/\/ebiztoday.news\/index.php\/2026\/04\/30\/kubernetes-control-plane-and-red-hat-drive-ai-infrastructure\/","title":{"rendered":"Kubernetes control plane and Red Hat drive AI infrastructure"},"content":{"rendered":"
\n

The recent KubeCon + CloudNativeCon Europe<\/span> event didn’t read like a celebration a lot as an admission. AI is in all places, however the systems underneath it are strained \u2014 not by models, but by layout.\u00a0<\/span>The information and machines that make AI are spread across clouds, edge sites and on-prem environments that never agreed on learn how to behave as one system \u2014 the core challenge driving the Kubernetes control plane for AI.<\/span><\/p>\n

Recent <\/span>research<\/span><\/a> shows that the vast majority of AI initiatives fail to succeed in production, with most breakdowns brought on by integration and operational execution challenges moderately than model performance.<\/span><\/p>\n

Paul Nashawaty<\/span><\/a>, principal analyst at theCUBE Research, describes the structural realities at the guts of the problem:\u00a0<\/span>\u201cAI is exposing a fundamental flaw in enterprise infrastructure; it was never designed to operate as a unified system. What KubeCon EU makes clear is that fragmentation across cloud, edge and on-prem is now the first barrier to production AI.\u201d<\/span><\/p>\n

That fragmentation now gets a reputation: sovereignty. Systems inherit policy, enterprise or regional borders. Those boundaries constrain where data and workloads can run, forcing AI systems to operate across distributed environments moderately than a single unified stack.\u00a0<\/span><\/p>\n

\u201cPossibly the finance BU has their Llama model; perhaps the accounting BU has an OpenAI model,\u201d said\u00a0Mike Barrett<\/a>, vp and general manager of Red Hat Hybrid Platforms at Red Hat Inc., who spoke with theCUBE, SiliconANGLE Media\u2019s livestreaming studio, during KubeCon EU. \u201c<\/span>What’s probably the most cost-effective strategy to surface the intelligence that you simply want back? Due to that, [Red Hat\u2019s enterprise customers] are on the lookout for a horizontal platform.\u201d<\/span><\/p>\n

Red Hat, the poster child of Kubernetes within the enterprise, goals to rein in fragmentation with a Kubernetes control plane for AI workloads across all environments.<\/span><\/p>\n

This feature is an element of SiliconANGLE Media\u2019s exploration of\u00a0how enterprises are constructing the control plane for AI, with Red Hat<\/span><\/span> playing a central role in shaping that approach.\u00a0<\/em>(* Disclosure below.)<\/em><\/p>\n

Kubernetes goes beyond orchestration to resolve AI challenges<\/h3>\n

Kubernetes was never designed for AI inference. It schedules containers. It doesn’t guarantee consistency across regions. That gap becomes visible when inference workloads move into production.<\/span><\/p>\n

\u201cThese models are doing an amount of compute that\u2019s hard to fathom, but after I refer to users of llm-d [an open-source, Red Hat-led Kubernetes-native inference framework hosted by CNCF, designed to scale distributed LLM workloads across clusters], they\u2019re not only attempting to construct a state-of-the-art performance system, they\u2019re also attempting to do these day-two operations,\u201d said\u00a0Robert Shaw<\/a> (pictured, left), director of engineering at Red Hat.<\/span><\/p>\n

That \u201cday-two\u201d problem is where AI systems often break \u2014 not in training, but in runtime behavior,\u00a0 latency swings, resource contention and policy drift. <\/span>Red Hat AI Enterprise<\/span> seeks to operationalize and speed up agentic AI and production inference with a unified, \u201cmetal-to-agent\u201d solution, in keeping with\u00a0<\/span>Jan Melen<\/a>, governing board vice chair on the Cloud Native Computing Foundation, which hosts Kubernetes and runs the KubeCon + CloudNativeCon conferences.<\/span><\/p>\n

\u201cCloud-native exists because of worldwide open-source collaboration model,\u201d \u00a0Melen said in the course of the KubeCon EU keynote<\/a>. \u201cHundreds of contributors from every region constructing shared infrastructure together.\u201d<\/span><\/p>\n

The implication will not be subtle. AI is pushing systems built on global consistency into environments defined by fragmentation.<\/span><\/p>\n

\u201cAgentic AI isn\u2019t a model problem \u2014 it\u2019s a platform architecture problem,\u201d said\u00a0Rob Strechay<\/a>, principal analyst at theCUBE Research. \u201cThe enterprises that win won\u2019t pick higher models; they\u2019ll construct higher infrastructure to run them.\u201d<\/span><\/p>\n

Kubernetes becomes less about orchestration and more about enforcing behavioral consistency across fractured environments, Strechay identified.<\/span><\/p>\n

Platform engineering makes Kubernetes the control plane for AI<\/h3>\n

While Kubernetes can unify control, it may\u2019t assume every team can operate that control directly. Enterprise adoption collapses when complexity is exposed raw.<\/span><\/p>\n

\u201cWhat we realized is that AI is being developed by data scientists, and as a part of that, they\u2019re constructing their very own infrastructure to run it on,\u201d said\u00a0<\/span>Brian Stevens<\/span><\/a> (pictured, right)<\/span>, senior VP and chief technology officer for AI at Red Hat.<\/span><\/p>\n

That gap between builders and operators is where platform engineering enters, in keeping with\u00a0<\/span>Strechay.<\/span><\/p>\n

\u201cFragmented tooling, skill gaps and operational complexity have gotten the true bottlenecks, driving a shift toward platform engineering and Kubernetes as a unifying control plane,\u201d he explained.<\/span><\/p>\n

The system stabilizes only when Kubernetes stops being exposed directly and becomes mediated through platforms that reduce friction. <\/span>Red Hat OpenShift AI<\/span> sits in that role, abstracting operational complexity into repeatable patterns, with model training, deployment, serving and inference for hybrid environments.<\/span><\/p>\n

Virtual machines extend Kubernetes as a substitute of resisting it<\/h3>\n

Enterprises don’t modernize every thing without delay. Billing systems and databases are likely to stay where they’re. Mainly, risk keeps legacy systems alive.<\/span><\/p>\n

Research<\/span><\/a> shows that 84% of IT decision-makers report difficulty managing separate VM and container environments, with siloed tools and fragmented operations driving inefficiency across hybrid infrastructure.\u00a0<\/span>If those VMs stay outside Kubernetes, the system stays split. But what if virtualization is brought into Kubernetes?<\/span><\/p>\n

\u201cWe predict virtualization and containers mustn’t live in silos; they ought to be on one platform \u2014 and KubeVirt makes that occur,\u201d said\u00a0<\/span>Daniel Messer<\/span><\/a>, senior manager of product management at Red Hat.\u00a0<\/span><\/p>\n

KubeVirt, a project now maturing at CNCF, extends Kubernetes into virtualization, allowing VMs and containers to share the identical control plane.\u00a0<\/span><\/p>\n

\u201cGraduating for us makes it more obvious for those that [KubeVirt] is deeply embedded within the Kubernetes ecosystem [and] the CNCF ecosystem,\u201d added\u00a0<\/span>Andrew Burden<\/span><\/a>, KubeVirt maintainer.<\/span><\/p>\n

The direction is consolidation of operational surfaces, not elimination of legacy systems.<\/span><\/p>\n

Sovereignty creates distribution, not centralization<\/h3>\n

Sovereign AI often looks like an answer, however it also imposes constraints. Laws block data from moving across borders. Policy blocks centralization. Enterprises split workloads across clouds, on-prem and edge environments whether their architecture is prepared or not, in keeping with\u00a0<\/span>Gabriele Bartolini of EnterpriseDB, who reframed the underlying principle in a <\/span>recent interview<\/span><\/a> with <\/span>The Recent Stack.<\/span><\/p>\n

\u201cTrue sovereignty starts with the database,\u201d Bartolini said. \u201cIn case your PostgreSQL isn\u2019t portable across environments, you don\u2019t really control your stack.\u201d\u00a0<\/span><\/p>\n

And he warns against assuming managed convenience equals control:\u00a0<\/span>\u201cConvenience is the cloud\u2019s biggest shortcut, but convenience isn\u2019t sovereignty. Real control means you possibly can move your database anywhere and it behaves the identical.\u201d<\/span><\/p>\n

Jan Melen\u2019s keynote draws a tough line contained in the sovereignty debate:\u00a0<\/span>\u201cWe should always separate code sovereignty from deployment sovereignty. The code itself stays global commons, shared, open, collaboratively developed.\u201d<\/span><\/p>\n

Deployment is where sovereignty bites. That’s where law and policy resolve where workloads can actually run, and under what conditions. T<\/span>he split is what Kubernetes tries to operationalize: global code, local execution.<\/span><\/p>\n

Ecosystems define whether the Kubernetes control plane for AI will work<\/h3>\n

No vendor can cover AI infrastructure alone. A Kubernetes control plane for AI only works if it spans systems as a substitute of replacing them. That puts the burden on the ecosystem \u2014 the shared standards, APIs and upstream projects that permit different tools operate as one system.<\/span><\/p>\n

Nashawaty points to Red Hat\u2019s role inside that upstream layer:\u00a0<\/span>\u201cRed Hat\u2019s influence extends well beyond its industrial platform. The corporate has long been one of the energetic contributors to the Cloud Native Computing Foundation ecosystem.\u201d<\/span><\/p>\n

That upstream work will not be cosmetic. It’s what keeps Kubernetes consistent across vendors. Without it, every distribution drifts and the control plane fractures into competing implementations.<\/span><\/p>\n

Except for contributing to open-source projects, Red Hat is partnering with corporations for scalable enterprise AI. Notably, <\/span>Red Hat AI Factory with Nvidia<\/span>\u00a0focuses on constructing, deploying and scaling AI infrastructure using Red Hat OpenShift and Nvidia accelerated computing for high-performance AI workloads.<\/span><\/p>\n

\u201cWhen as many as 75% of enterprises report double-digit AI failure rates tied to fragmented systems, it\u2019s clear the bottleneck has shifted to infrastructure,\u201d said Nashawaty, underscoring the associated fee when upstream infrastructure coordination breaks down, especially in the case of AI.
<\/span><\/p>\n

That failure rate will not be about missing features. It reflects systems that can’t operate together. Ecosystems prevent Kubernetes from collapsing into one other silo \u2014 the precise consequence it’s purported to avoid.\u00a0\u00a0<\/span><\/p>\n

Kubernetes becomes the production layer for AI<\/h3>\n

AI doesn’t collapse infrastructure in a single place. It stresses every seam without delay. Kubernetes becomes the layer that attempts to carry those seams together.<\/span><\/p>\n

Stevens described the shift toward consolidating fragmented systems onto a single platform:\u00a0<\/span>\u201cIt\u2019s a really powerful concept to reapply that \u2014 and likewise consolidate it on the identical platform with fewer vendors and fewer attack surface for changes in numerous learning curves, which I feel has been the ability of Kubernetes all along.\u201d<\/span><\/p>\n

That consolidation only works if the ecosystem holds. Melen underscores what happens if it doesn\u2019t:\u00a0<\/span>\u201cIf sovereignty results in fragmentation, we risk undermining the trillions of dollars of value that open source has already brought globally.\u201d<\/span><\/p>\n

The system doesn’t change into simpler, however it becomes governable.\u00a0<\/span>Kubernetes isn\u2019t the tool of selection since it is ideal. The industry chooses it because fragmentation leaves no alternative coordination layer that scales.<\/span><\/p>\n

Red Hat\u2019s bet is that abstraction through Kubernetes is the one viable strategy to keep AI operational across disparate worlds.<\/span><\/p>\n

(* Disclosure: TheCUBE is a paid media partner for the KubeCon + CloudNativeCon NA event. Neither Red Hat Inc., the first sponsor of theCUBE\u2019s event coverage, nor other sponsors have editorial control over content on theCUBE or SiliconANGLE.)<\/em><\/p>\n

Photo: SiliconANGLE<\/h5>\n
\n
\n

Support our mission to maintain content open and free by engaging with theCUBE community. Join theCUBE\u2019s Alumni Trust Network<\/strong>, where technology leaders connect, share intelligence and create opportunities.<\/p>\n