Choosing storage for a VMware estate is a different exercise than choosing storage in general. It’s not just about IOPS and capacity, it’s about how well the array integrates with vSphere, VCF, and the operational tooling your team already relies on: vVols support, VASA providers, SRM compatibility, snapshot and replication behaviour under load, and how the platform behaves when Storage vMotion or Distributed Resource Scheduler is moving workloads around underneath it.
The VMware landscape itself has also been evolving. Broadcom’s ownership of VMware has changed licensing models and pushed some organisations to evaluate alternatives, but the vSphere/VCF ecosystem remains the default for most enterprise virtual estates, and vendor engineering investment in VMware integration hasn’t slowed. If anything, the arrays below have doubled down on tighter API-level integration precisely because that’s still where the workloads are.
Below are five storage platforms worth shortlisting for a VMware environment in 2026, based on breadth of VMware integration, real-world deployment maturity, and fit across different workload profiles. This isn’t a ranked “best to worst” list, the right choice depends heavily on your existing hardware relationships, team skillset, and whether you’re optimising for cost, performance, or operational simplicity.
HPE Alletra Storage MP
HPE’s Alletra Storage MP line (the modern successor to the Nimble and Alletra 6000/9000 generations) is built around a disaggregated architecture that lets you scale compute and capacity independently, which matters more than it sounds for VMware clusters where workload profiles shift over time. It carries forward InfoSight predictive analytics, which has a solid track record of flagging storage-related issues before they cause a support call, and it integrates with vSphere through VASA providers for vVols alongside native VMware plugins.
Where it fits well: organisations already invested in the HPE ecosystem, or those wanting a single management plane across storage, servers, and increasingly, private cloud via HPE’s Morpheus platform.
Consideration: as with any disaggregated architecture, the value compounds over a multi-year refresh cycle, it’s less compelling as a one-off, standalone purchase.
Pure Storage FlashArray//X
Pure’s FlashArray//X remains one of the most widely deployed all-flash platforms in VMware estates, largely on the strength of its Purity operating environment and the Evergreen upgrade model, which allows controller refreshes without the forklift replacement cycle that older SAN architectures required. VMware integration runs deep, native vVols support, VASA 3.0, and a plugin that surfaces array-level detail directly inside vCenter, which is genuinely useful when you’re troubleshooting a datastore-level performance issue without switching consoles.
Where it fits well: teams that prioritise operational simplicity and predictable non-disruptive upgrades over granular manual tuning.
Consideration: Purity’s management model is somewhat opinionated by design, teams wanting deep, hands-on control over storage internals may find NetApp or HPE’s toolsets offer more configurability.
NetApp AFF and ASA A-Series
NetApp’s ONTAP-based AFF and ASA arrays are a strong fit for organisations running mixed VMware and non-VMware workloads side by side, since ONTAP’s unified block and file capabilities mean the same array can serve vSphere datastores and general-purpose file shares without a second platform. The VMware plugin (VSC – Virtual Storage Console) integrates provisioning, cloning, and snapshot-based backup directly into vCenter, and ONTAP’s FlexClone technology is genuinely useful for VDI and test/dev environments where you need to spin up near-instant, space-efficient VM clones.
Where it fits well: environments with diverse storage protocols (NFS, iSCSI, FC, SMB) sitting alongside VMware, or teams that want a single storage OS across primary and secondary use cases.
Consideration: ONTAP has a steeper learning curve than some competitors, the flexibility comes with more configuration surface area to understand.
Dell PowerStore
PowerStore has matured into a credible modern midrange option, with its AppsON capability allowing certain applications to run directly on the array controllers alongside VM storage, a distinctive approach in this category. VMware integration is comprehensive: vVols via VASA, VCF workload domain support, and management through vRealize Operations and vRealize Orchestrator for teams already standardised on VMware’s broader management stack.
Where it fits well: organisations wanting broad protocol and ecosystem compatibility, or those already running significant Dell server infrastructure alongside their virtual estate.
Consideration: the AppsON model is a genuinely different architectural choice, so it’s worth a proper technical evaluation against your specific application requirements rather than assuming it behaves like a standard array.
VMware vSAN (as part of VCF)
It would be an odd omission to leave VMware’s own storage layer off a list of storage options for VMware environments. vSAN takes a fundamentally different approach to the four platforms above, rather than a dedicated array sitting outside the compute layer, it aggregates local storage across the hosts in a vSphere cluster into a single distributed datastore. For organisations already committed to VMware Cloud Foundation, vSAN removes a layer of architecture entirely, and recent releases have continued to push scale-out performance and features like Auto-RAID for storage policy management.
Where it fits well: greenfield VCF deployments, edge and ROBO sites where a dedicated SAN isn’t practical, and organisations wanting to reduce the number of moving parts in their infrastructure stack.
Consideration: vSAN ties your storage lifecycle to your compute lifecycle and to VMware’s own licensing model, worth weighing carefully given the licensing changes since the Broadcom acquisition, particularly for organisations wanting more architectural independence.
It’s not just about the array
Everything above treats storage as a standalone decision, which is a reasonable way to compare platforms on technical merit, but it’s worth stepping back if what you’re planning is a broader infrastructure refresh rather than a storage-only swap-out. Several of the vendors in this list, HPE and Dell in particular, sell a lot more than arrays, and if compute, storage, and virtualisation are all up for renewal around the same time, evaluating them together can change the shape of the decision considerably.
HPE’s relevant offering here is Private Cloud Business Edition (PCBE), more recently rebranded as HPE Private Cloud PC3000. It’s a factory-integrated stack combining ProLiant compute, Alletra Storage MP, and a choice of virtualisation, either VMware or HPE’s own Morpheus VM Essentials software, with support for running both side by side during a phased migration. The pitch is a single management plane covering VM lifecycle, infrastructure provisioning, and firmware, rather than separately procuring and integrating compute, storage, and virtualisation layers. For organisations weighing up their virtualisation licensing options in the current landscape, having that flexibility built into the hardware stack from day one is a genuinely useful property, regardless of which hypervisor a team ultimately runs.
Dell has a broadly comparable proposition with VxRail and its APEX consumption models, bundling compute and storage into a single validated, VMware-integrated appliance with a unified support and lifecycle model.
None of this changes the storage-specific comparison above; PowerStore and Alletra still stand on their own technical merits as arrays. But if the conversation in your organisation is really “we need to refresh our virtual estate” rather than “we need new storage,” it’s worth having that broader conversation with your vendor(s) before the storage decision gets made in isolation.
How to choose
Rather than picking from a list, the useful exercise is working backwards from your actual constraints:
- Protocol requirements – do you need unified block and file, or is block-only sufficient?
- Team skillset – is your team already fluent in one of these platforms, or is this a genuine reset?
- Refresh cycle appetite – do you want non-disruptive, evergreen-style upgrades, or are periodic hardware refreshes acceptable?
- Workload mix – pure VMware, or VMware alongside Kubernetes, AI/ML pipelines, or unstructured data growth?
- Scope of the refresh – is this storage-only, or does it make sense to evaluate compute, storage, and virtualisation together, as with HPE PCBE or Dell VxRail/APEX?
- Licensing and cost model changes – particularly relevant given how VMware’s own commercial terms have shifted; this affects the total cost picture more than the storage line item alone.
In practice, the solution that fits your environment tends to emerge once these constraints are laid out clearly, rarely is it obvious from a spec sheet alone. If you’re weighing up options for a VMware refresh, whether that’s a like-for-like storage swap or a full infrastructure reset, a proper assessment of your current environment, growth trajectory, and operational preferences will get you to the right answer faster than a straight vendor comparison.
