All-Flash Arrays
All-Flash Storage Arrays
An all-flash array is bought for one of two reasons: a latency floor a hybrid array cannot reach, or the estate consolidation that comes from putting several ageing arrays onto one platform with inline reduction. These cover both, and they are not interchangeable — the difference is mostly protocol, data services and how you intend to grow.
Pure FlashArray spans mission-critical block, a high-end consolidation platform and a QLC capacity tier. Dell PowerStore covers all-NVMe block and a unified block-and-file variant. NetApp runs from unified AFF platforms to the SAN-optimised ASA line. Which one fits usually follows from the protocols you already run and the data services your applications and backup vendor expect.
Effective-capacity figures published for these platforms are the manufacturers' own, quoted at their stated reduction ratios. Your ratio depends on your data, and databases that are already compressed will not reach the datasheet numbers. We would rather set that expectation now than have you discover it after commissioning.
10 configurations below · quotes returned within 48 business hours · purchase orders accepted
All-Flash Storage Arrays we configure and supply
Every system below is quoted to your workload — accelerator count, CPU, memory, storage and fabric are specified together rather than sold as a fixed SKU.
FlashArray//X R4 (all-NVMe TLC, mission-critical block)
Low-latency all-flash block storage for your most demanding databases.
- Predictable low latency
- Higher effective capacity
- Non-disruptive lifecycle
FlashArray//XL170 R5 (sub-150µs high-end consolidation)
Top-tier all-flash performance for your most data-intensive applications.
- Headroom to consolidate
- Extreme low latency
- Resilient by design
FlashArray//C90 R4 (QLC capacity-optimized all-flash)
All-flash economics at petabyte scale for tier-2 and capacity workloads.
- Disk economics, flash benefits
- Dense petabyte capacity
- Unified operations
Dell PowerStore 3200T All-NVMe Block Array (NVMe/FC)
Active/active all-NVMe block storage with end-to-end NVMe/FC
- End-to-end NVMe/FC
- Always-on resilience
- Efficient effective capacity
Dell PowerStore 1200T All-NVMe Unified Array (NVMe/TCP)
Right-sized all-NVMe unified storage for block and file consolidation
- Unified block and file
- Ethernet-native NVMe
- Scale with expansion shelves
NetApp AFF A90 All-Flash Unified Array (200GbE, ONTAP)
High-end unified flash that feeds AI pipelines without starving GPUs.
- GPUs stay fed
- One platform, every protocol
- Six-nines availability
NetApp AFF A800 All-Flash NVMe Array (100GbE, ONTAP AI)
Proven NVMe storage behind NVIDIA DGX deep-learning environments.
- Validated AI architecture
- Throughput at scale
- Data managed, not copied
NetApp ASA A400 SAN-Optimized All-Flash Array (32Gb FC, NVMe/FC)
Symmetric active-active SAN flash for databases that cannot blink.
- Always-on transactions
- Predictable low latency
- Lower fabric overhead
NetApp AFF A1K Modular Flash Flagship (64Gb FC, 400GbE)
Flagship modular flash that consolidates the data center at scale.
- Consolidate at scale
- Top-tier performance
- Future-proof fabrics
Synology FlashStation FS6400 (24-bay all-flash, dual Xeon Silver, 10GbE)
All-flash 2U storage built for I/O-intensive databases and virtualization.
- Consistent low latency
- Scales with demand
- Delivered ready to run
What we need to quote accurately
A configuration quote takes minutes when these are known and days of back-and-forth when they are not. You do not need all of them to start — send what you have.
- The workload: model sizes, training or inference, and expected concurrency
- Node count, and whether this is a single system or a scaling cluster
- Rack power and cooling available per rack, and inlet temperature
- Existing estate — the vendor and management tooling you already run
- Network fabric: InfiniBand, Ethernet, and the speed you are standardised on
- Timeline, and whether the budget is approved or being built
How we quote
- 1
Send the requirement
An email, a bill of materials, or a rough description of the workload. All three work.
- 2
We validate the configuration
Accelerator, chassis, fabric, power and cooling checked against each other before anything is priced.
- 3
Itemised quote within 48 hours
Line-by-line pricing, current availability and lead time, with warranty terms stated.
all-flash storage array — buyer questions
FlashArray, PowerStore or NetApp AFF — how do we choose?
Largely by what you already run and which protocols you need. Pure is block-first with a very simple operating model. PowerStore is the natural fit inside a Dell estate and gives block and file on one appliance in the unified variant. NetApp is the answer when ONTAP data services, replication or a genuinely multiprotocol namespace matter. If you have no incumbent, it usually comes down to the data services your applications and backup vendor expect.
NVMe/FC or NVMe/TCP?
NVMe/FC if you already have a Fibre Channel fabric and the switching lifecycle to support it — it remains the lower-jitter option for latency-sensitive databases. NVMe/TCP if you are consolidating onto Ethernet and would rather not maintain two fabrics. NVMe/TCP is now credible for most workloads that used to demand FC, but it puts more weight on how well your Ethernet fabric is configured for lossless behaviour.
Are the published effective capacities realistic?
They are manufacturer figures at a stated reduction ratio. Virtual machine estates and file data often exceed those ratios; already-compressed databases, encrypted volumes and media files fall well short. We size on raw capacity plus a reduction assumption you are comfortable defending internally, and we say so on the quote rather than presenting the best case as the number you are buying.
Do we need a six-nines platform?
Only if you can articulate what an hour of downtime costs. Symmetric active-active controllers and non-disruptive upgrades genuinely earn a high availability specification. But for most estates, a dual active-active array with a tested replication target and a rehearsed failover procedure gets you further than a higher availability figure on a single site. We quote the replication and the second site as line items so you can see that trade-off in cash.
Compare with other platforms
AI & HPC Storage
Storage for AI Training Clusters
View systemsEnterprise NVMe
Enterprise NVMe SSDs
View systemsDell PowerEdge
Dell PowerEdge Rack & Tower Servers
View systemsHPE ProLiant
HPE ProLiant Rack & Tower Servers
View systemsBrowse the full range of GPU servers and AI clusters, or the component catalogue for drives, memory, optics and spares.