Dynamic Resource Allocation
Kubernetes Wasn’t Built for GPUs. Make It Behave
Kubernetes counts whole GPUs and treats pods as disposable. An LLM pod is neither. Share the silicon with MIG/MPS/time-slicing and stop paying for idle ...
Sneha Gullapalli | | A100, AI infrastructure, AI Workloads, cloud native AI, Dynamic Resource Allocation, GPU autoscaling, GPU cost reduction, GPU optimization, GPU partitioning, GPU sharing, GPU time-slicing, GPU utilization, H100, Karpenter, KServe, Kubernetes DRA, Kubernetes GPU scheduling, LLM Inference, model caching, multi-instance GPU, NVIDIA GPU Operator, NVIDIA MIG, NVIDIA MPS, scale-to-zero, VRAM
Stop Treating GPUs Like Web Pods
Kubernetes schedules accelerators as opaque integers, and your bill pays for it. Share the silicon, scale on the right signal and keep weights out of the image ...
Veera Ravindra Divi | | AI infrastructure, AI serving, autoscaling, cloud costs, cloud native AI, DCGM exporter, DRA, Dynamic Resource Allocation, GPU costs, GPU scheduling, GPU sharing, GPU utilization, GPUs, inference workloads, KEDA, kubernetes, Kubernetes GPU scheduling, LLM Inference, MIG, model weights, MPS, NVIDIA GPUs, NVIDIA MIG, Prometheus, scale-to-zero, time-slicing
Kubernetes v1.36 Promotes Stability, Compatibility & Reproducibility
Kubernetes v1.36 (Spring 2026) introduces 70 enhancements, including major security hardening for the Kubelet API and the debut of Workload-Aware Scheduling (WAS) for AI/ML. This release focuses on fine-grained resource health, stable ...
Adrian Bridgwater | | AI/ML Infrastructure, CI/CD, cloud native security, cloud-native applications, Cluster Hardening, container security, containers, CSI Token Redaction, developers, Distributed Training, DRA, Dynamic Resource Allocation, External Token Signing, Gang Scheduling, K8s v1.36, Kubelet API Authorization, kubernetes, Kubernetes Enhancements 2026., Kubernetes v1.36, microservices, Node Logs, open source, PodGroup API, Resource Health Status, storage, Volume Group Snapshots, WAS, workload-aware scheduling
What to Expect From Kubernetes 1.36
Kubernetes 1.36 launches April 22, 2026, marking a major shift in networking as Ingress-Nginx retires in favor of the more scalable Gateway API. Key updates include bolstered Linux User Namespaces for better ...
Adrian Bridgwater | | admission control config, CloudNativeCon, cluster security, container isolation, Deployment Stability, DRA, Dynamic Resource Allocation, EKS, fat image anti-pattern, gateway API, ingress-nginx retirement, Karpenter, KubeCon Europe, Kubernetes 1.36, Linux user namespaces, LLM weights, manifest-based admission control, OCI artifacts, platform engineering, security patches, specialized hardware, taints and tolerations, upgrade risk, VolumeSource, WatchCache
The Efficiency Era: How Kubernetes v1.35 Finally Solves the “Restart” Headache
Kubernetes v1.35 introduces in-place resource resizing, revolutionizing how stateful workloads are managed. Discover the benefits of dynamic resource allocation, traffic distribution, and the improvements that enhance operational efficiency for platform engineers ...
Pavan Madduri | | AI/ML workloads, cloud costs, Dynamic Resource Allocation, efficiency era, FinOps, immutability, Kubernetes architecture, Kubernetes enhancements, Kubernetes v1.35, Openshift, operational efficiency, resource resizing, self-healing infrastructure, Stateful Workloads, system performance, traffic distribution, vertical scaling
Kubernetes v1.35 Arrived, Right On Workload-Aware Schedule
Discover the latest enhancements in Kubernetes workload scheduling, including the Workload API and gang scheduling features aimed at optimizing application performance and management ...
Adrian Bridgwater | | autoscaling, cloud-native applications, Dynamic Resource Allocation, Gang Scheduling, kubernetes, Kubernetes v1.35, Multi-Node Scheduling, Opportunistic Batching, Performance Optimization, Pod Management, resource allocation, Scheduling Algorithms., Scheduling Improvements, Scheduling Latency, software engineering, Workload API, Workload Scheduling

