Dell PowerScale AI Storage
Dell PowerScale NAS storage offers an end-to-end data solution to feed GPU-intensive servers and accelerate AI workflows at scale - the ideal partner for systems such as 3XS EGX, MGX and HGX servers and NVIDIA DGX AI appliances. The OneFS NAS operating system provides enhanced performance with massive scalability to support the most demanding file-based workloads, while enabling you to reduce hardware footprint and optimise storage resources. OneFS software can handle any unstructured data and deploy anywhere – in the datacentre, at the edge, core or in the cloud.
Performance to Suit your Demands
Dell PowerScale NAS offers flexibility in platform choice with the all flash F-series, hybrid H-series and archive-centric A-series systems.
TLC & QLC Flash Systems
F-series TLC- and QLC-based AI appliances, built on Dell's OneFS parallel file system support the largest, most generative AI workloads, such as LLMs and deep learning recommender systems.
Hybrid Systems
H-series hybrid systems offer better cost-efficiency than all flash-based solutions. However, they still support a variety of large-scale AI workloads.
Archive Disk
A-series archive data systems remain essential too, especially in the context of AI applications, and needs to be accessible. Dell systems handle both active and cold archive workloads at the lowest total cost.
PowerScale AI Cluster Configurations
To meet the requirements of a variety of workloads, the Dell PowerScale OneFS architecture leverages a choice of modular systems, offering up to 2.9PB per 2U node using 24 122TB NVMe SSDs. Scaling from three to 252 nodes delivers massive scalability topping out at 186PB raw capacity or up to 737PB when using compression and deduplication features - all connected via 200Gb/s InfiniBand or Ethernet networking.
The Dell PowerScale AI Range
PowerScale systems can drive workloads at the edge, in the datacentre or in the cloud. These nodes are available in a cost-effective modular form factor which deliver outstanding performance to address the needs of all AI workloads.
| System | F-series | H-series | A-series |
|---|---|---|---|
| DRIVE BAYS (per 2U) | Up to 24x 2.5in | Up to 60x 2.5in | Up to 60x 2.5in |
| DRIVES SUPPORTED | 3.84TB, 7.68TB, 15.36TB, 30TB, 62TB, 122TB NVMe | 2TB, 4TB, 8TB, 250TB, 12TB, 16TB, 20TB, 24TB SATA / SAS | 2TB, 4TB, 8TB, 250TB, 12TB, 16TB, 20TB, 24TB SATA / SAS |
| NETWORKING | 2x InfiniBand HDR / 200GbE | 2x InfiniBand EDR / 100GbE | 2x InfiniBand EDR / 100GbE |
Ready to Deploy Your AI Infrastructure?
As a leading NVIDIA Elite Partner and the UK's only NVIDIA-certified DGX Managed Services Partner, Scan is ideally suited to design and deploy your GPU-accelerated compute & DDN-based storage cluster. Contact our AI experts for further information on 01204 474210 or on [email protected]