Seagate Exos AP2U24

Overview

The Exos AP2U24 is fundamentally different from the SCALE family of JBOD expansion enclosures — it's a genuine storage server, with dual x86 compute nodes built directly into the chassis alongside 24 bays of 2.5" storage. Where SCALE enclosures rely entirely on a separate host server for processing, the AP2U24 brings its own compute to the storage, making it a self-contained platform rather than an expansion unit.

Each of the two controllers is powered by an AMD EPYC processor, giving the AP2U24 genuine server-grade processing power alongside its storage capacity — useful for workloads that benefit from compute sitting physically close to the data it's working with.

Key Specifications

Why Choose the AP2U24

The AP2U24's defining advantage is combining compute and storage in one box. For workloads where you'd otherwise need a separate compute server talking to a storage array over the network, having both in a single 2U chassis reduces both latency and complexity — data processing happens right next to where the data lives.

It's also built to be right-sized rather than over-specified: with a choice of 8, 12 or 16-core AMD EPYC configurations, you can match compute capacity to your actual workload rather than paying for headroom you don't need. And because it shares its design and field-replaceable components with the rest of the Exos AP and SCALE ecosystem, servicing and spares are simplified across a mixed deployment.

When you need more raw capacity than the 24 internal bays provide, the AP2U24 can expand out to further SCALE enclosures — up to 8 chassis in total — without needing a separate host server to manage that expansion.

Common Use Cases

• High-bandwidth, low-latency workloads — Applications where compute needs to sit close to storage rather than communicate over a network hop.
• Medical imaging and surveillance — Environments processing continuous image or video sequences alongside local storage.
• Unified storage platforms — Running software-defined storage stacks (e.g. file, block and object management) directly on the storage server itself.
• Small to mid-sized deployments — Organisations wanting server-grade compute and dense storage without managing separate compute and storage tiers.

Further Reading