The Microsoft’s Azure Container Networking team is announcing Retina, a cloud-native container networking observability platform that enables Kubernetes users, admins, and developers to visualize, observe, debug, and analyze Kubernetes’ workload traffic irrespective of Container Network…
You can now choose from preset LLMs with images hosted by AKS and split inferencing across multiple lower-GPU count VMs
You can now better understand the infra costs associated with running applications at the namespace and cluster levels and identify opportunities to optimize resource utilization through an Azure native experience.
You can now schedule workloads in Fleet based on cost and availability of resource heuristics.
YOu can now use Gen 2 VM SKUs for Windows nodepools in AKS.
You can now run GPU workloads, such as machine learning, video encoding, large simulations, and gaming, on Windows nodepools in AKS.