We are excited to announce that Mirantis k0rdent AI has been validated in the NVIDIA-Certified Hypervisors program. The k0rdent AI virtualization stack has been certified through the NVIDIA-Certified Hypervisors program on the NVIDIA GB200 NVL72 validation track. GPU virtualization can help reduce the traditional performance trade-off. What you get: • Predictable, near bare metal performance • GPU allocation matched to each workload, full node or sub-node • Strong tenant isolation with the VM as the tenancy boundary • One control plane for VMs and containers Learn more about running virtualized GPU workloads on k0rdent AI. buff.ly/dAEofSx #Mirantis #k0rdentAI #NVIDIA #GPUVirtualization @nvidia

AI builders rely on a cloud that moves at their pace. Today, IREN takes another step forward with the completion of our acquisition of Mirantis. By bringing together IREN’s owned and operated data centers and compute with Mirantis’ flexible, interoperable software layer, we’re giving customers greater choice and control in how they deploy and scale AI workloads. Developers and enterprises want flexibility as the AI landscape evolves. We're building an open foundation that grows with you, so you’re ready for what's next. Read more: iren.gcs-web.com/static-file…

k0s is now Certified Kubernetes AI Conformant. 🎉 Big milestone for the project: the CNCF has certified k0s under its Certified Kubernetes AI Conformance program at Kubernetes v1.35. It's a more demanding bar than standard Kubernetes conformance, and it gives independent, public verification that the AI and ML workloads organizations run today (model training, inference serving, and AI agents) operate reliably on k0s using standard, upstream tooling. The certification was validated on a live cluster: k0s running on NVIDIA T4 GPUs on Azure. k0s met every mandatory capability, along with optional ones such as GPU time-slicing. A lean, single-binary distribution often raises the question of what has been sacrificed to keep it small. For AI workloads, the answer is nothing of consequence. Any project that runs on upstream Kubernetes runs on k0s, with no proprietary components to adopt. For teams evaluating where to run training and inference, this offers independent, reproducible verification rather than a vendor claim. Read the full breakdown: buff.ly/nfJNAcs

Clusters deployed by k0rdent are now Certified Kubernetes AI Conformant. 🎉 Here's the question every platform team asks. If we standardize our clusters on one tool, are we quietly giving something up when it's time to run AI? The CNCF just answered it for k0rdent. Its clusters passed Certified Kubernetes AI Conformance at v1.35, validated on a live cluster with NVIDIA T4 GPUs. And it's not a datasheet claim, but independent, public proof that anyone can rerun. What it means if you run AI on k0rdent: • No lock-in. The clusters are standard Kubernetes, so your existing tooling for GPU scheduling, traffic routing, and monitoring just works. Nothing proprietary to adopt. • Every cluster, verified. k0rdent's job is managing many clusters at once. Now each one it provisions is a verified, standardized place to run AI, so your teams don't relearn a bespoke environment every time. k0rdent covered all mandatory capabilities plus optional ones like sharing a single GPU across workloads. Read the full breakdown: buff.ly/LMstn14

Everyone's racing to reinvent the stack for AI. Our CEO's take: you don't have to. AI doesn't change the fundamentals of compute. You still need IaaS. You still need Kubernetes. What changed is that the layer everyone abandoned when they ran to the hyperscalers is suddenly the one that matters most. Mirantis never abandoned it. We kept the OpenStack business and kept building IaaS while it was unfashionable, which is exactly why Nvidia is close to us now. In this clip, Alex Freedland, CEO & Co-founder of Mirantis, explains why the AI era rewards the companies that understood infrastructure all along. Watch the full interview: buff.ly/Gln4Hk5

What if your AI infrastructure could be operational the same day the hardware arrives? That's exactly what Mirantis, NVIDIA, and Dell are delivering for enterprises right now. Built on a composable architecture that works with your existing environment, Mirantis k0rdent AI removes the barriers between investment and outcome. Here's what that looks like in practice: • Speed: GPUs operate the same day hardware arrives, with Metal-to-Model™ deployment in days, not months • Validated partnerships: As an inaugural NVIDIA AI Cloud-Ready ISV partner, fully validated on Dell PowerEdge GPU servers • True flexibility: No rip-and-replace, no proprietary lock-in, and the freedom to evolve as your business changes Backed by a decade of enterprise cloud expertise, Mirantis brings the operational knowledge to turn your AI investment into real business outcomes. See how it works: buff.ly/9tu40M6 #NeoCloud #EnterpriseAI #SovereignAI #AIInfrastructure

The question with multi-agent systems in regulated environments is not whether agents can act. It is how far their authority extends, and what stops them from exceeding it. Autonomy creep is a hard problem, especially when agents can spawn and orchestrate other agents. William Rizzo, Global Field CTO at Mirantis, will take the stage with Francesco Beltramini, Field CTO from ControlPlane at OSFF London to present a pragmatic, cloud native approach to enforcing agent guardrails on real-world Kubernetes stacks. 📅 25 June | 12:00-12:35 BST | OSFF London Catch it live: buff.ly/OYE5Uad

Your AI team shouldn't wait weeks to get a model running. Watch how Mirantis k0rdent AI and Run:ai spin up a live LLM inference workload in minutes, with intelligent GPU scheduling and secure image delivery built in automatically. Faster access to GPUs. Less infrastructure overhead. More time on what actually matters. Explore the demo: buff.ly/ANxrvWH 🇫🇷 We'll be at RAISE Summit in Paris on July 8-9. Book a meeting with our team to see it live. buff.ly/vWWhc3o #RAISESummit #NeoCloud #EnterpriseAI #AIInfrastructure #AICloud #SovereignAI #k0rdentAI #AIFactory

Enterprise platform teams are stretched thin, managing hybrid infrastructure sprawl amid fragmented tools, inconsistent governance, and a lack of unified visibility across the estate. Mirantis k0rdent AI gives you one control plane to: • Manage Kubernetes fleets across on-prem, cloud, and edge • Enforce consistent governance and compliance fleet-wide • Operationalize AI workloads with GPU provisioning and MLOps integration, without creating new silos Get the product overview: buff.ly/hueXeEP 💡 See it in action on June 24th. Join Martin Stadler, Field CTO for AI Infrastructure, for a live webinar on building a Private Enterprise AI Factory on infrastructure you control. buff.ly/RdbVjyI #PrivateAI #EnterpriseAI #AIInfrastructure #AIFactory

The CAPEX impact of underutilized on-prem compute is significant. In regulated environments, a large portion of that idle capacity is by design reserved for compliance headroom, workload isolation, and governance requirements. 5-Spot is built to reclaim it. RBC Capital Markets' first open-source contribution to FINOS, 5-Spot brings time-based compute scheduling to Kubernetes using Cluster API patterns, helping organizations get more value from the infrastructure they already own. Erick Bourgeois, Head of Kubernetes Platform Engineering at RBC Capital Markets @rbccm, will lead a hands-on workshop at OSFF London on exactly this, co-sponsored by Mirantis and @controlplaneio . 📅 24 June | 9AM to 1PM BST | OSFF London If you're based in London, this one's worth your morning. Save your seat: buff.ly/aUNPe4s

What does it actually look like to manage 50+ edge Kubernetes clusters without growing your ops team? In this session, we break it down into three principles: • Standardize: One declarative template as the source of truth. Drift gets eliminated at the source, not patched after an incident. • Centralize: One control plane for every cluster across every edge site. Unified observability, policy enforcement, and lifecycle management. • Automate: Self-healing clusters that reconcile continuously, so your engineers aren't getting paged at 2am. Watch the full on-demand session for the complete framework. buff.ly/MAn8WT2 #Kubernetes #EdgeComputing #EdgeKubernetes #MultiCloud #HybridCloud #MultiCluster #PlatformEngineering

What does it actually look like to go from bare infrastructure to a running LLM in production? We built an interactive demo to show exactly that. Watch Mirantis k0rdent AI deploy GPU and multi-cloud clusters, configure NVIDIA Run:ai workloads, and spin up a live NVIDIA NIM inference service, all through a self-service experience that requires no manual setup. This is what Metal to Model looks like in practice. buff.ly/E1SX5iM 🇫🇷 Heading to RAISE Summit in Paris? See a live version at Booth #29a. Book a meeting with our team in advance to go deeper. buff.ly/4BQNyEQ #RAISESummit #NeoCloud #EnterpriseAI #AIInfrastructure #AICloud #SovereignAI #k0rdentAI #AIFactory

Your Kubernetes fleet may be running at 30% utilization, and in financial services, that's often by design. 5-Spot, RBC Capital Markets' first open-source contribution to FINOS, is built to change that. It transforms Cluster API into a spot-aware scheduling substrate, helping organizations reclaim idle capacity without compromising cluster health, compliance requirements, or operational guardrails. Join Erick Bourgeois, Head of Kubernetes Platform Engineering at RBC Capital Markets, on June 24 for a hands-on workshop where you'll learn how to: • Deploy 5-Spot on a live k0smotron-managed cluster • Identify and classify spot-eligible workloads • Explore the architectural decisions behind running spot capacity safely in regulated environments If you're a platform engineer, cloud architect, or SRE responsible for Kubernetes at scale, especially in regulated industries, this session is worth your morning. Special thanks to ControlPlane, Erick Bourgeois, and RBC Capital Markets for helping bring this workshop to the community. Save your seat: buff.ly/aUNPe4s

Everyone is teaching AI to write code. Nobody is teaching it to run the systems that code lives on. That gap is the biggest open opportunity in infrastructure right now. @randybias (VP of Tech @MirantisIT ) on episode 3 of The Root Cause. Live now: piped.video/GdfuNtM1WTM

AI Factories are becoming the gold standard for delivering modern AI and ML services. But building one that is sovereign, multi-tenant, and GPU-powered at scale? That is where most teams hit a wall. Mirantis k0rdent AI provides a proven blueprint for operators and service providers to rapidly launch differentiated GPU cloud services. It ensures full data sovereignty, secure multi-tenant isolation, and flexible scalability across hybrid and edge environments. Download our AI Factory Reference Architecture to get the technical and business insights you need to build and operate a next-generation AI cloud: buff.ly/qVLnqX4 🇫🇷 Attending RAISE Summit in Paris? Book a meeting with our team and see how it all comes together. buff.ly/DMYuYIz #RAISESummit #NeoCloud #EnterpriseAI #AIInfrastructure #AICloud #SovereignAI #k0rdentAI #AIFactory

Most people think scaling AI means getting more GPUs. That's only part of the answer. Running AI at scale means agentic pipelines, databases, and traditional compute all working alongside your GPUs. When GPU demand goes up, everything underneath it goes up too. The entire stack moves together. This is why the companies that understand the full stack, not just the GPU layer, are the ones that will define this era. Mirantis has been building that stack for 15 years. In this clip, Alex Freedland, CEO & Co-founder of Mirantis, breaks down what an AI workload actually looks like, and why betting on GPUs alone is only half the story. Watch the full interview: buff.ly/VmYXMU1

The AI Executive Salon started as a private gathering at GTC. Now it's one of the most sought-after evenings in the AI infrastructure world. Paris is next! 🇫🇷 On July 8, founders, operators, investors, and technology leaders shaping the future of AI compute will come together on a rooftop above Paris, hosted by Mirantis, @IREN_Ltd, and @nvidia. If you're attending RAISE Summit, this is the evening you don't want to miss. 🗓 July 8 | 7:00 PM CET 📍 ROOF at Madame Rêve, Paris Spots are limited and curated. Apply now: buff.ly/M30K3GF #RAISESummit #AIExecutiveSalon #AIInfrastructure #AICommunity