Dell PowerScale F910 All-Flash Scale-Out NAS

The Dell EMC PowerScale F910 is a purpose-built all-flash scale-out NAS node designed as the successor to the F710, delivering substantially higher node bandwidth and storage density through PCIe Gen 5 NVMe drive technology. Built on the proven OneFS operating system, the F910 integrates into existing PowerScale clusters seamlessly, enabling enterprises to scale capacity and performance non-disruptively from a handful of nodes to thousands while maintaining a single global namespace. The architecture supports inline data reduction technologies including compression and deduplication, ensuring maximum effective capacity utilization without sacrificing the low-latency access profiles demanded by modern GPU-accelerated AI and machine learning pipelines.

The F910 targets environments where throughput-intensive workloads such as large-scale AI model training, inference data staging, genomics analysis, media and entertainment post-production, and financial analytics require sustained sequential and random I/O at extreme bandwidth levels. With dramatically increased per-node bandwidth compared to its predecessor, a single F910 cluster can serve concurrent data streams to hundreds of GPU nodes or AI compute clusters simultaneously, eliminating storage as a bottleneck in accelerated computing infrastructures. The node ships with high-speed network interconnects optimized for 100GbE and 200GbE fabric connectivity, ensuring that NVMe-tier performance reaches compute nodes at wire speed.

Administered through the Dell PowerScale InsightIQ analytics platform and the OneFS WebUI or CLI, the F910 provides granular performance visibility, multi-protocol support spanning NFS, SMB, S3, and HDFS, and enterprise-grade data protection through SmartProtect replication and SyncIQ policy-based data management. Multi-tenancy, role-based access control, and compliance capabilities make the F910 equally suited to regulated industries requiring strict data governance alongside high-performance throughput. The result is a converged performance-and-capacity platform that eliminates the traditional trade-off between flash density and scalability in unstructured data environments.

Ideal for

• AI and machine learning training dataset staging: serving large NVMe-backed datasets at high throughput to GPU compute clusters running frameworks such as PyTorch or TensorFlow, minimizing GPU idle time during data ingestion phases
• Genomics and life sciences research: supporting parallel read-intensive workloads from next-generation sequencing pipelines where hundreds of concurrent analysis jobs require sustained high-bandwidth access to petabyte-scale reference and sample datasets
• Media and entertainment post-production: providing shared all-flash storage for 4K and 8K video editing, color grading, and visual effects rendering workflows requiring deterministic low-latency access across large creative teams
• High-frequency financial analytics: delivering sub-millisecond access to tick data, risk model datasets, and time-series archives that require simultaneous read access from multiple quantitative analysis compute nodes
• Enterprise backup and archival acceleration: acting as a high-speed landing zone for backup data streams from large virtualized environments before tiering to object or tape, enabling faster recovery time objectives through rapid restore from all-flash tier
• Multi-cloud data hub: serving as an on-premises high-performance NAS anchor integrated with Dell APEX cloud storage services, enabling hybrid workflows where AI training occurs on-premises and inference or overflow compute scales to public cloud

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.