Everpure's Portworx Enhances Kubernetes Storage for AI and VMs, Streamlining Hybrid Workloads

Everpure has announced significant updates to its Portworx platform, focusing on improving storage management for AI workloads and virtual machines (VMs) running on Kubernetes. Key features include pre-staged volumes for AI and machine learning environments, the general availability of Kube Datastore (KDS) for VMs on Kubernetes, and enhanced observability within the Portworx operations center. These updates aim to reduce data access delays for AI applications and provide a more consistent approach to storage management across diverse infrastructure. This development is particularly important for organizations that are increasingly looking to consolidate their compute environments onto Kubernetes. By addressing the specific storage needs of AI and VMs, Portworx helps platform engineers overcome common challenges like slow data loading for AI models and the operational overhead of managing VM storage separately. The ability to manage these disparate workloads from a single Kubernetes-native platform offers substantial operational efficiencies and reduces complexity. The move aligns with a broader industry trend towards Kubernetes as a universal control plane for various workload types, including traditional virtualized applications and emerging AI/ML pipelines. As enterprises continue their digital transformation journeys, the need for a unified data management layer that can handle the unique demands of each workload type becomes critical. This trend is further evidenced by the continuous evolution of Container Storage Interface (CSI) drivers and the increasing maturity of Kubernetes for stateful applications. In practice, these updates mean that DevOps and AI teams can expect faster startup times for AI training and inference jobs due to data being pre-positioned closer to compute resources. For IT operations teams, Kube Datastore offers a more streamlined way to migrate and manage virtual machines within Kubernetes, leveraging familiar storage controls while benefiting from Kubernetes' orchestration capabilities. Practitioners should evaluate how these features can be integrated into their existing Kubernetes deployments to optimize resource utilization, reduce operational friction, and accelerate the delivery of AI-driven applications and modernized VM estates. The trade-off might involve initial configuration efforts to define appropriate storage policies for pre-staging and KDS integration, but the long-term benefits in performance and simplified management are likely to outweigh these.