Applications
Combine Ustorage host cards, switch boards, retimers and chassis into purpose-built Gen6 systems. Below are the scenarios our customers deploy most — each with a recommended starting configuration.
Scenario 01 — AI Computing
Large models need many GPUs talking to each other without bottlenecks. A Gen6 expansion system lets you add GPU capacity outside the server while keeping bandwidth and link stability where training runs need them.
✓ Scale GPU count beyond the server chassis
✓ Stable Gen6 links for multi-day training runs
✓ One validated stack instead of mixed vendors
Recommended configuration
Request this configurationScenario 02 — Test & Validation
Device makers and integrators need a reference-grade fabric to qualify Gen6 SSDs, GPUs, cables and firmware. Repeatable topologies and clean signal paths make compliance and regression testing faster and more trustworthy.
✓ Repeatable, documented topologies
✓ Clean signal paths with Gen6 retimers
✓ Switch management for flexible lane partitioning
Recommended configuration
Request this configurationScenario 03 — Storage
Attach dense NVMe storage over Gen6 with predictable latency. Ideal for data pipelines, media workflows and caching tiers that must never stall.
✓ More drives without a bigger server
✓ Predictable latency over longer cable runs
✓ Easy to grow as datasets grow
Recommended configuration
Request this configurationScenario 04 — HPC
Engineering simulation, rendering and scientific computing benefit from GPU capacity that can grow without replacing servers. A double-chassis Gen6 platform keeps large GPU pools organised and serviceable.
✓ Grow GPU capacity in steps
✓ Serviceable, rack-friendly design
✓ Same building blocks as our AI configurations
Recommended configuration
Request this configurationDescribe your application and target quantities — our engineers will propose a configuration and quotation.