Scale-out all-flash for AI/ML and HPC — 1,200 TB per blade, 400GbE, NFS/S3/SMB
The FlashBlade//S R2 is Pure Storage's scale-out all-flash platform for unstructured data — delivering up to 1,200 TB per blade and 50% higher performance than the previous generation for AI/ML training, HPC, genomics, and high-performance analytics. A single unified platform for NFS, S3, and SMB at petabyte scale.
Get Pricing — Speak to Our Team
Competitive pricing · Response within 4 hours
FlashBlade//S R2 blades deliver 20–25% better multidimensional performance over competitors for RAG, AI training, inference, and simulation workloads. The modular architecture scales capacity and performance independently — increasing blades adds throughput and IOPS, adding chassis adds raw capacity — enabling exact-fit scaling for AI infrastructure without stranded resources.
Up to 16 × 400GbE uplinks in a multi-chassis FlashBlade//S configuration provide the network bandwidth to feed GPU clusters at training throughput. Integrated networking is upgradeable via the eXpansion Fabric Module (XFM) — no separate switching infrastructure required for chassis-to-chassis connectivity.
FlashBlade//S supports billions of files and objects in a single system through Purity//FB's distributed metadata architecture. Native NFS (v3 and v4.1), SMB 3.x, and S3 APIs are served from the same unified platform — eliminating the protocol silos that force organisations to maintain separate file and object storage systems.
Zero Move Tiering is Pure Storage's revolutionary approach to data placement in heterogeneous all-flash clusters — enabling intelligent, granular file-level data access across capacity tiers without physically moving data. This provides performance predictability, infrastructure agility, and cost efficiency that traditional tiering (which moves cold data to slower media) cannot achieve.
Rapid Replicas synchronise file metadata across geographically distributed FlashBlade sites every 5 minutes — enabling on-demand file access at remote locations without full dataset replication. Only accessed data transfers, reducing bandwidth consumption for EDA, AI/ML, genomics, and other distributed compute workloads.
The Evergreen//Forever subscription makes FlashBlade//S the last scale-out storage system an organisation needs to buy. It includes latest-generation blade upgrades every 3 years (Ever Modern), trade-in credit on blade model upgrades (Ever Agile), and future DirectFlash Module density consolidations — all without downtime or data migrations.
The Pure Storage FlashBlade//S R2 is the definitive scale-out all-flash platform for unstructured data — purpose-built for the most demanding AI/ML training, HPC, genomics, EDA, and data analytics workloads. With next-generation S200 R2 and S500 R2 blades, FlashBlade//S R2 delivers up to 50% higher performance than the previous generation and up to 1,200 TB per blade.
The modular architecture of FlashBlade//S enables independent scaling of compute and storage. A minimum configuration of 7 blades in a single chassis grows up to 10 blades per chassis and up to 10 chassis in a multi-chassis configuration — with each XFM (eXpansion Fabric Module) providing the chassis interconnect. Up to 16 × 400GbE uplinks feed the output of a fully scaled FlashBlade//S deployment.
Purity//FB OS manages all system resources — blades, DirectFlash Modules, and networking — at a global level. This global media management unlocks 20% more usable capacity from NAND compared to systems using commodity SSDs. Native NFS (v3 and v4.1), SMB 3.x, and S3 protocol support handles billions of files and objects simultaneously without protocol gateway overhead.
Zero Move Tiering provides intelligent, policy-driven data access across heterogeneous blade configurations without physically moving data — maintaining predictable performance while optimising cost efficiency. Rapid Replicas enable near-instant file metadata synchronisation across geographically distributed sites, supporting global collaboration workflows for distributed AI, genomics, and EDA teams.
FlashBlade//S R2 is purpose-built for GPU-scale AI infrastructure — providing the sustained throughput required to feed NVIDIA H100 clusters during LLM pre-training, fine-tuning, and RAG workloads, with 20–25% better performance over competing solutions in these specific workload types.
The ability to support billions of files in a single system and scale NFS throughput independently of capacity makes FlashBlade//S the reference platform for genomics sequencing pipelines, protein structure prediction, and clinical data analysis — where sequential read performance at petabyte scale is the primary storage requirement.
FlashBlade//S provides high-throughput backup targets and rapid restore capabilities for enterprise backup workflows — with SafeMode Snapshots protecting backup data from ransomware and fast-read architecture enabling full restores in a fraction of the time required by disk-based recovery systems.
Trusted UK IT reseller since 2003 — supplying businesses of all sizes with genuine, competitively priced technology.
Servnet is an authorised IT reseller with direct access to leading product lines at competitive pricing.
Our engineers will help you select the right product and advise on compatibility with your existing infrastructure.
Most products ship from UK stock. Express delivery available for urgent deployments.
All products are genuine, brand-new, with full manufacturer warranty and RMA support.
We can pre-configure your hardware before shipping — reducing on-site deployment time and cost.
Post-sale support from our certified engineers. We're here if anything goes wrong.
All-flash at disk economics — starts at 4.5 PB, NFS/S3/SMB, scales to 10 chassis
Top-tier enterprise NVMe storage — 150µs latency, 9.4 PB effective, 45 GB/s
Mission-critical NVMe storage — 250µs latency, 4.4 PB effective, 99.9999% availability
