Skip to content
Enterprise Storage | Pure Storage Evergreen | Intelligent Visibility
Infrastructure / Enterprise Storage

Storage that gets better over time — without forklift upgrades

Pure Storage FlashArray with Evergreen delivers sub-millisecond performance, 5:1 data reduction, and a subscription model that eliminates the disruptive 3–5 year refresh cycle that defines traditional enterprise storage.
Explore the AIM Program Talk to an Architect
<1ms
Latency
5:1
Data Reduction
99.9999%
Availability
0
Forklift Upgrades
Pure Evergreen

The storage ownership model that should have always existed

End the refresh cycle

Traditional enterprise storage follows a predictable and painful pattern: buy an array, run it for 3–5 years, then rip it out and replace it with the next generation. Every refresh means downtime, data migration, re-qualification, and a new capital expenditure conversation.

Pure Evergreen eliminates this cycle. Controllers are upgraded non-disruptively — your data stays where it is, your applications never go offline, and your storage gets faster and denser over time through the subscription rather than through forklift replacements.

This isn't just a financial model change. It's an operational one. Your storage team stops managing refresh projects and starts managing capacity and performance.

  • Non-disruptive controller upgrades — no downtime, no data migration
  • Performance improves over time as new controllers are deployed
  • Capacity added on demand without replacing the array
  • Software updates and new features delivered continuously
  • Predictable costs — no surprise capital requests every few years
Recommended

Evergreen//Flex

Subscription with ownership. Combine the predictability of a subscription with the flexibility to add reserve capacity on demand. Non-disruptive upgrades included. Scale storage independently from compute without capital planning cycles.

Evergreen//Forever

CapEx ownership with non-disruptive upgrades. You own the hardware, and Pure upgrades the controllers in place as part of the support agreement. The traditional buy model, but without the forklift refresh.

Evergreen//One

Full storage-as-a-service. Consumption-based pricing with guaranteed SLAs for performance, capacity, and availability. Pure manages the infrastructure — you consume storage like a cloud service, on-prem.

The Shift

What changes when you stop buying arrays and start subscribing to outcomes

Traditional storagePure Evergreen
Refresh cycleForklift replacement every 3–5 years with downtime and data migrationNon-disruptive controller upgrades — storage improves in place
Capacity planningBuy for projected peak 3 years out and hope you sized correctlyAdd capacity on demand as workloads grow — no over-provisioning
Budget modelLarge CapEx spikes every refresh cycle, hard to predictPredictable subscription — finance loves it, procurement plans for it
Performance over timeDegrades as array ages and approaches end of lifeImproves as new controllers are deployed through the subscription
Software updatesManual, often deferred due to risk and maintenance windowsContinuous, non-disruptive updates with new features included
Operational burdenRefresh projects consume engineering time for monthsNo refresh projects — team focuses on capacity and performance
Platform

Pure Storage FlashArray — the performance foundation

FlashArray is the hardware platform that Evergreen runs on. All-flash, NVMe-native, with inline data reduction that typically delivers 5:1 efficiency. It's designed for the workloads where latency and throughput directly impact business performance.

Sub-millisecond latency

Consistent, predictable performance for latency-sensitive workloads. Databases, transactional applications, and real-time analytics that can't tolerate I/O variability.

5:1 average data reduction

Inline deduplication and compression that's always on with no performance penalty. A 50TB raw array typically delivers 250TB+ of effective capacity.

NVMe end-to-end

NVMe-native architecture from the host interface through the internal bus. No protocol translation bottlenecks — the entire I/O path is designed for flash.

Active-active controllers

Dual controllers operating simultaneously for both performance and resilience. No active-passive waste — both controllers serve I/O at all times.

Space-efficient snapshots

Near-instant, metadata-only snapshots that consume no additional capacity until data changes. Protect data without the storage tax of traditional snapshot approaches.

Built-in replication

Synchronous and asynchronous replication for DR and business continuity. Replicate to another FlashArray on-prem or to Pure Cloud Block Store in AWS.

Storage Networking

The fabric between compute and storage matters

FlashArray performance is only as good as the network connecting it to your compute nodes. We design storage networking on Arista deep-buffer switches that protect high-throughput storage traffic and eliminate the need for a separate Fibre Channel SAN.

Ethernet-based storage protocols

Run NVMe-oF, iSCSI, or NFS on the same Ethernet fabric as your compute traffic. One unified network instead of separate LAN and SAN infrastructures — lower cost, simpler operations.

NVMe-oF (RoCEv2) NVMe/TCP iSCSI NFS
NVMe-oF Protocol Comparison

Arista deep-buffer switching

7280-class switches designed to handle the incast congestion patterns that throttle storage traffic on standard switches. Lossless Ethernet with proper buffer allocation for sustained high-throughput I/O.

Arista 7280 Deep Buffers Lossless Ethernet PFC / ECN
FC vs Ethernet Comparison
Cloud Extension

Your storage operating model, extended to AWS

Pure Cloud Block Store runs the same Purity operating environment natively on AWS. Same data services, same management experience, same replication — in the cloud. No application refactoring required.

Pure Cloud Block Store

Run Pure storage in AWS with the same Purity//FA software you run on-prem. Replicate snapshots to the cloud for DR, spin up dev/test environments from production data, or burst storage capacity during peak demand.

  • Same management experience on-prem and in AWS
  • Replication from FlashArray to CBS for DR
  • Dev/test from production snapshots in cloud
  • Data mobility without application changes
Cloud Storage Continuity Guide

Hybrid cloud with AIM

Within the AIM fabric, Pure Cloud Block Store works alongside Nutanix NC2 on AWS and AWS EVS to create a complete hybrid continuity model — compute, storage, and virtualization with parity across on-prem and cloud.

  • Storage continuity via Pure CBS
  • Compute continuity via NC2 on AWS
  • VMware continuity via AWS EVS
  • Aegis co-managed across all environments
Aegis co-managed storage operations: IVI manages your Pure environment through Aegis — monitoring array health and performance, managing capacity growth, coordinating controller upgrades, and handling replication configuration. Your team retains control and visibility while we handle the operational lifecycle.

Design your storage architecture with Evergreen in mind

We'll assess your current storage environment, capacity trajectory, and performance requirements — then design a Pure Storage architecture with the Evergreen model that fits your operational and financial goals.

Explore the AIM Program Talk to an Architect

Other

6 resources

Other Resources