Strategic PartnersIssue 102 records
Everpure
Portworx by Everpure used KubeCon (Nov 10) to announce pre-staged volumes that copy model weights to Kubernetes nodes ahead of time so AI inference pods start and recover faster, alongside GA of Kube Datastore for VMs on FlashArray and an IDC MarketScape Leader rating. The AI feature is roadmap for Q1 2027, and independent StorageReview coverage notes Portworx has published no start-up or failover figures and names only FlashArray as a backend. No SLED deployment is cited; for state, campus and research-computing platform teams it is a VMware-exit and on-prem inference planning signal to validate, not proof.
- Evidence records
- 2
- Cross-source patterns
- 1
- Evidence classes
- 1 vendor claim1 independent reporting
- Outcomes
- 2 emerging
- Source freshness
- 2 new this fortnight
- Research completed
- 2026-11-11
Choose a role to see its takeaway beside every record in the ledger.
Patterns across the evidence
Synthesis · Lighthouse Advisory interpretation
Portworx's AI inference feature arrives as roadmap with no published start-up figures
Everpure's Nov 10 announcement presents pre-staged volumes as faster AI start-up and recovery, and StorageReview's same-day coverage confirms the feature is expected Q1 2027, may change in scope and timing, and has no published start-up figures; the GA part of the release is VM storage (KDS) on FlashArray.
Operating questionBefore counting AI start-up gains in a Kubernetes or VMware-exit plan, can we measure pod time-to-ready and failover in our own cluster once the feature ships?
Supporting evidenceEverpure (Portworx by Everpure)StorageReview
Evidence ledger
Full record · every source keeps its link and limitations
Portworx Unveils Advanced AI and VM Capabilities on Kubernetes; Everpure Named a Leader in 2026 IDC MarketScape Report
Everpure's Nov 10, 2026 KubeCon announcement for Portworx (its Kubernetes data platform) lists four items: pre-staged volumes that distribute warm node-local copies of model weights and other volume data so AI inference pods start and recover faster (expected Q1 2027, not yet available); general availability of Kube Datastore (KDS) for VMs on Kubernetes on supported SAN arrays; launch-partner support for Red Hat OpenShift 5; and operations-center observability plus bring-your-own-backup. It also quotes IDC's 2026 MarketScape for Container Data and Infrastructure Management Software naming Everpure a Leader and calling Portworx a good fit for mission-critical VMs, containers or AI workloads on Kubernetes, while noting Everpure is still building virtualization feature parity. The companion Nov 9 Portworx technical blog says a pipeline hydrates one source volume with a versioned model bundle from an approved repository, copies are confirmed ready before pods are scheduled, and model updates can use a blue/green rollout.
Why it matters, evidence and limitations
- Why it matters
- Moderate and indirect. State IT shops, research-computing centers and university platform teams moving off VMware or hosting on-prem AI inference on Kubernetes are the plausible audience; no SLED customer is named. Why now: announced at KubeCon NA on Nov 9-10 and not previously archived.
- Evidence and measured results
- Vendor product description and a vendor-quoted analyst rating; no measurements, sample, baseline or customer outcome.
- Limitations and uncertainty
- Vendor announcement; no startup-time, recovery or cost measurements, no customer or public-sector deployment, and the AI feature is roadmap (Q1 2027, scope and timing subject to change). The IDC MarketScape is quoted by the vendor; the full IDC report is paywalled and was not read.
Portworx Kube Datastore Reaches GA With Single-Replica Pools on FlashArray
StorageReview (Harold Fritts, Nov 10, 2026) independently reports the same release: KDS is GA in Portworx Enterprise 3.7.2, using single-replica 'dynamic pools' on a shared array that leave data protection to the array, with FlashArray the only named backend. It relays Portworx's figure that shared-nothing replication carries roughly 2x-3x storage overhead and 3x-5x network I/O amplification, and notes Portworx has not published failover times. For AI, it describes pre-staged volumes as targeting inference start-up delays when tens to hundreds of gigabytes of model weights are re-downloaded after a pod moves, and states that Portworx expects the feature in Q1 2027, says scope and timing may change, and has not published startup-time figures.
Why it matters, evidence and limitations
- Why it matters
- Moderate: gives SLED buyers the questions the vendor post leaves open (one named backend, no failover or start-up numbers). Why now: same-day independent coverage of the Nov 10 KubeCon release, not previously archived.
- Evidence and measured results
- Trade-press reporting of vendor briefing with explicit notes on unpublished metrics; no testing.
- Limitations and uncertainty
- Trade-press news write-up based on the vendor briefing; it adds scrutiny (missing figures, single named backend) but no independent testing. The 2x-3x and 3x-5x overhead figures are Portworx's, not measured by StorageReview. No public-sector content.
How to read this edition
Source findings, measured results and limitations come from the cited publications. Patterns, operating questions, role takeaways and implementation considerations are Lighthouse Advisory interpretation, stated as questions to validate locally rather than guaranteed outcomes. Vendor and operator claims are labeled as claims. Full research method.
- Vendor claim
- A supplier-provided assertion that has not been upgraded to independent evidence.
- Independent reporting
- Independent reporting with attributable sources but without a formal evaluation design.