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 of Kube Datastore on supported SAN arrays. Everpure also said it was named a Leader in the 2026 IDC MarketScape for Container Data and Infrastructure Management Software.
- Pre-staged volumes are expected in Q1 2027.
- Kube Datastore is generally available on supported SAN arrays.
- Portworx is available today for Red Hat OpenShift 5 deployments.
- Operations Center adds visibility into storage and virtualization operations.
- Everpure was named a Leader in the 2026 IDC MarketScape report.
Pre-staged volumes target AI startup delays
Portworx’s pre-staged volumes are designed to reduce the time AI and machine learning workloads spend waiting for data. The company said large model weights may need to be downloaded repeatedly before an application can start, creating storage-related delays.
Expected in Q1 2027, the capability will allow platform teams to distribute warm local copies of persistent volume data across cluster nodes. Storage policies will help workloads choose an appropriate local copy when they start or move to another node.
According to Everpure, this approach is intended to support faster startup and recovery in AI environments where workloads are frequently scheduled, moved or restarted. The announcement does not provide performance figures.
Kube Datastore becomes generally available
Portworx has also made Kube Datastore, or KDS, generally available on supported SAN arrays. The service is aimed at enterprises moving virtual machines into Kubernetes while continuing to run containerized and AI workloads on the same platform.
KDS supports dynamic and static pools, providing storage placement and capacity controls that build on the performance and resilience of native arrays. Everpure said this can allow workloads to be configured with fewer replicas.
Platform engineers can carry out operations such as rebalancing and maintenance through Kubernetes native workflows, while VM administrators get familiar datastore concepts and workflows, including storage mobility. Everpure describes this as a way to make the move to Kubernetes more intuitive and to reduce the need for separate storage procedures.
Broader platform support and industry recognition
Portworx is available today for customers deploying Red Hat OpenShift 5, where Everpure is a launch partner. The company has also expanded Operations Center observability with visibility into pool inventory, live migration progress and hyperconvergence state.
In addition, Portworx is expanding bring-your-own-backup support for third-party backup products alongside its own backup and disaster recovery capabilities. The goal is to give teams more flexibility in protecting workloads across Kubernetes environments.
Everpure said the IDC MarketScape recognized it as a Leader in its 2026 report. The report says Portworx is a good fit for enterprises running mission-critical VMs, containerized applications or AI workloads on Kubernetes across multiple distributions or mixed storage, particularly those seeking integrated backup and disaster recovery.
Greg Muscarella, General Manager, Portworx, at Everpure, said: “Portworx gives teams a consistent approach to running containers and VMs on Kubernetes and managing their data across public clouds and on-premises infrastructure.” The updates position Portworx as a common data management layer for organizations operating different workload types on one Kubernetes platform.
