What Is 2N Redundancy?_
In a 2N redundant system, every power distribution component—from utility feeds and UPS modules to PDUs and rack-level power strips—is duplicated. If one entire path fails (e.g., a UPS goes offline or a feeder breaker trips), the remaining path instantly supplies 100% of the required power with no single point of failure. This design is common in high-availability data centers but doubles the capital cost and floor space for electrical gear.
Technical Details
2N redundancy requires two separate electrical rooms, each with its own UPS, battery banks, and distribution panels, feeding two independent power whips to each rack or row. The rack’s power supplies must be configured in an active/active or active/standby mode, with each PSU connected to a different A-side or B-side feed. Transfer times are effectively zero because both paths are live simultaneously; no static transfer switch is needed at the rack level. Compliance with the relevant standard (e.g., TIA-942’s Tier IV or Uptime Institute’s Fault Tolerant class) typically mandates 2N for both power and cooling paths.
How Leviathan Systems Works with 2N Redundancy
When we deploy a Leviathan Systems GPU cluster (a hypothetical example), the customer’s facility must provide two independent power drops per rack—labeled A and B—so that our field crew can land both feeds on the dual-input PDU before racking compute nodes. We verify that each GPU node’s power supplies are split across the A and B buses during commissioning, and we label every cable and breaker to prevent accidental cross-feeding.
Related service
GPU Rack Assembly & Rack-and-Stack →