Kubernetes

CoreWeave Joins CNCF as a Platinum Member
CoreWeave has become a Platinum member of the Cloud Native Computing Foundation, expanding its support for Kubernetes and the AI-native cloud community. The company…

NVIDIA Donates Agent Substrate Runtime to CNCF Sandbox
NVIDIA has donated its open-source Agent Substrate runtime to the Cloud Native Computing Foundation Sandbox, targeting the challenges of running large numbers of…

NVIDIA Snapshot restores GPU workloads in seconds on Kubernetes
NVIDIA has introduced Snapshot, an open-source checkpoint and restore system for GPU workloads on Kubernetes that can bring initialized inference replicas online in…

Portworx adds AI startup and VM storage features for Kubernetes
Portworx by Everpure has unveiled new capabilities for running AI workloads and virtual machines on Kubernetes, including pre-staged volumes and the general availability…

Datadog adds Kubernetes inference workload monitoring in Preview
Datadog has introduced its Kubernetes inference workload feature in Preview, giving engineers a consolidated view of model configuration, serving components, pod state…

KubeRay Joins CNCF, Bringing the PARK AI Stack Under Linux Foundation Governance
KubeRay is joining the Cloud Native Computing Foundation, according to Anyscale, placing the open source Kubernetes operator for Ray under Linux Foundation governance.…

Datadog expands MCP Server for Kubernetes and process investigations
Datadog has expanded its MCP Server with tools for Kubernetes control plane health, Live Processes search, and a processes skill that helps AI agents choose the right…

Red Hat OpenShift 5 adds agentic operations and longer support
Red Hat OpenShift 5 is now generally available, bringing agentic workflows for cluster operations, a revised support lifecycle and additional security controls to the…
For developersKubeflow Trainer v2.4 Adds Workload-Aware Scheduling for TrainJob
Kubeflow Trainer v2.4 adds Workload-Aware Scheduling to TrainJob, allowing distributed AI and machine learning workloads to be scheduled as coordinated groups rather…

CoreWeave adds guided self-service to its SUNK AI training system
CoreWeave is extending SUNK, its Slurm on Kubernetes training system, with a guided self-service preview for deploying and managing clusters through the CoreWeave…

Kinference: Buoyant Releases Open Source Inference Control Plane for vLLM
Buoyant has released Kinference, an open source control plane for building inference platforms on vLLM and Kubernetes. According to the company, its KV-cache-aware load…

NVIDIA AI Cluster Runtime Targets Day-One Kubernetes Support for Vera Rubin NVL72
NVIDIA has detailed how its AI Cluster Runtime (AICR) packages tuned Kubernetes configurations for Vera Rubin NVL72 systems, with the aim of having environments ready on…

NVIDIA Workload Map Lets Kubernetes Tools Read AI Workloads Without Custom Adapters
NVIDIA has introduced Workload Map, a common description of Kubernetes workload structure that lets schedulers, controllers and platforms interpret distributed AI…

Darktrace 7.2 brings behavioral AI security to Kubernetes runtime activity
Darktrace has released version 7.2, which allows Darktrace / HYBRID NETWORK to extend its behavioral security into Kubernetes environments. According to the company, it…

NVIDIA AI Cluster Runtime Packages Dynamo Into Validated Kubernetes Recipes
NVIDIA has detailed how its AI Cluster Runtime (AICR) packages the Dynamo distributed inference framework and its supporting infrastructure into validated recipes for…

Daytona2 targets the agent runtime gap that Kubernetes leaves open
Daytona says teams running AI agents on Kubernetes end up building their own execution layer, and it is pitching Daytona2 as a purpose-built sandbox runtime that fills…
Products and toolsNew Relic Infrastructure 360 maps cloud dependencies and config changes to cut MTTR
New Relic has introduced Infrastructure 360, an advanced layer on its observability platform that maps cloud dependencies and tracks configuration changes so SREs and…

NVIDIA Mission Control 2.5 Brings Autonomous Recovery to Kubernetes
NVIDIA Mission Control 2.5 extends its autonomous recovery engine to Kubernetes-based workloads managed with NVIDIA Run:ai. Until now, the engine, which detects…
Chips and infrastructureHow Netflix keeps GPUs healthy for weeks-long training and inference jobs
Netflix has described how it detects failing GPUs and network hardware, replaces bad machines automatically and lets long training jobs survive hardware loss. The…

Microsoft's Agent Lightning v1.0 Trains Agents With Their Real Harnesses
Microsoft Research Asia has open-sourced Agent Lightning v1.0, a rebuilt reinforcement learning framework of about 3,500 lines of code that lets the same harness used in…

Linux Foundation: 94% of European Firms Say Digital Sovereignty Matters
Digital sovereignty is now firmly part of European technology strategy, according to the 2026 State of Open Source in Europe report from the Linux Foundation. The study…

NVIDIA AICR v1.0 locks down GPU Kubernetes cluster configuration
NVIDIA has released version 1.0 of AI Cluster Runtime (AICR), a set of version-locked, validated recipes for configuring GPU-accelerated Kubernetes clusters. According…

CoreWeave's SUNK runs Slurm inside Kubernetes on one shared compute pool
CoreWeave has introduced SUNK ("SlUrm oN Kubernetes"), an implementation that runs the Slurm workload manager on top of Kubernetes so that HPC and AI teams can use a…