Seagate Exos EX5U84

Overview

The Exos EX5U84 is the high-density flagship of the SCALE family, packing 84 individual 3.5" drive bays into a single 5U enclosure. Like the EX2U12 and EX2U24, it's a pure SAS-attached JBOD expansion unit rather than a self-contained RAID array — it connects directly to a host server or existing storage controller, adding raw drive capacity without a separate controller layer to manage.

Where it differs is scale. This is the enclosure to reach for when you're planning storage infrastructure in the multi-petabyte to exabyte range, where minimising rack units per drive matters as much as the drives themselves.

Key Specifications

Why Choose the EX5U84

The EX5U84's case is simple: fewer rack units per drive. Where a 2U12 or 2U24 enclosure needs multiple chassis to reach serious capacity, a single EX5U84 already holds 84 drives, and chains further with additional 5U84 units up to a maximum of 336 drives — making it the natural choice once you're scaling past what 2U enclosures can efficiently deliver.

High availability is built in throughout: dual cooling, dual power/cooling modules, dual I/O modules, and dual data paths to every drive, so no single component failure takes the enclosure offline. A drawer-style chassis design gives engineers direct access to drives for maintenance without disturbing the rest of the array, and every major component — drive carriers, controllers, power supplies, fan modules and expander cards — is hot-swappable.

One practical detail worth knowing at the planning stage: drive quantities need to be in multiples of 14, and a system supports a maximum of two different drive sizes at once — both straightforward to design around, but worth factoring into your initial specification.

Common Use Cases

• Exabyte-scale archive and backup — Organisations building out long-term retention infrastructure where density per rack unit is critical.
• Large-scale media and content storage — High-capacity unstructured data, video archives and content libraries.
• Data lakes and analytics platforms — Bulk capacity behind compute clusters processing large datasets.
• Hyperscale and service provider environments — Where floor space and power efficiency per terabyte are primary planning constraints.

Further Reading