What is Tier IV?

Tier IV is the most demanding classification in the Uptime Institute's Tier Standard for datacenter design, construction, and operations. First introduced in the 1990s and formally codified in Uptime Institute publication Tier Standard: Topology, a Tier IV facility must sustain any single equipment failure or planned maintenance event without impacting critical load. This is achieved through fully redundant, concurrently maintainable infrastructure across all paths: power, cooling, and network.

The defining requirement at Tier IV is 2N+1 redundancy, meaning every critical component (UPS modules, generators, chillers, pumps) has two fully independent capacity systems plus an additional redundant unit. Additionally, all distribution paths must be dual and physically separate so that no single point of failure exists. The facility must also tolerate a fault without human intervention; it must self-isolate and reroute around the failed component. This is called fault tolerant operation. Planned maintenance, such as swapping a generator or testing a switchboard, can be performed without any shutdown or risk to the load because the redundant parallel path carries the full load during work.

Tier IV sits at the top of the four-tier Uptime Institute hierarchy. At the low end, Tier I has no redundancy. Tier II adds partial redundancy. Tier III adds concurrent maintainability but still tolerates only a limited set of failures. Tier IV is the only tier that achieves full fault tolerance. Typical uptime for a Tier IV facility is 99.995% annually, allowing no more than about 26 minutes of downtime per year. Because of the immense capital cost (often 2-3 times that of a Tier III build), Tier IV is reserved for mission-critical workloads: financial trading systems, cloud hyperscaler core regions, government emergency response, and airline reservation systems.