PDU Types Explained: Basic, Metered, Managed and Switched

A Power Distribution Unit (PDU) is a device that distributes electrical power to equipment within a server rack or data centre cabinet. At its simplest, a PDU is a rack-mountable power strip that takes a single power input and distributes it across multiple outlets. At its most advanced, a PDU is a network-connected intelligent device that monitors power consumption per outlet, allows remote power cycling of individual devices, provides environmental sensor data, and integrates with data centre infrastructure management (DCIM) platforms. For IT resellers, understanding PDU types is essential because the right PDU choice directly impacts a customer's ability to manage power capacity, ensure uptime, and control energy costs.

A basic PDU is the entry-level option — essentially a rack-mountable power board with no intelligence. It takes a single power input (typically a standard Australian 10A or 15A plug, or an IEC C20 inlet for higher-capacity units) and distributes it across multiple outlets. Basic PDUs have no monitoring, no network connectivity, and no remote management capabilities. They may include a circuit breaker for overload protection and surge suppression, but otherwise they are passive devices. Basic PDUs are appropriate for small server rooms, network closets, and environments where power monitoring is not required — typically edge deployments or cost-sensitive installations where the priority is simply getting power to equipment safely.

A metered PDU adds power monitoring to the basic distribution function. Metered PDUs include a local display (typically a small LCD or LED panel on the unit itself) and often network connectivity via Ethernet, allowing administrators to monitor total power consumption remotely via a web interface, SNMP, or the vendor's management platform. Some metered PDUs provide only inlet-level metering (total load on the entire PDU), while more capable models offer per-outlet metering that shows exactly how much power each connected device is drawing. Per-outlet metering is invaluable for capacity planning — it allows administrators to see which outlets are underutilised and where additional load can be safely added without exceeding the circuit capacity.

Metered PDUs are the sweet spot for many deployments. They provide the visibility needed to manage power capacity and identify overloaded circuits before they trip, without the additional cost and complexity of switched outlets. They are especially useful in colocation environments where power billing is based on actual consumption — the metered PDU provides the data needed to verify invoices and optimise power usage. Most metered PDUs also support threshold alerting, sending SNMP traps or email notifications when power draw exceeds configured warning and critical levels.

A switched PDU (sometimes called a managed PDU) includes everything a metered PDU offers plus the ability to remotely switch individual outlets on or off. This capability is transformative for remote site management — if a server hangs and cannot be rebooted via software, an administrator can remotely power-cycle the specific outlet feeding that server without affecting any other equipment in the rack. This eliminates the need to dispatch a technician to physically pull a power cable, saving time, money, and reducing downtime from hours to minutes.

Switched PDUs also support outlet-level power sequencing — turning outlets on in a defined order with configurable delays between each. This is critical during power restoration after an outage: you want storage arrays to come online before servers, and servers before network switches, to prevent boot failures and data corruption. Advanced switched PDUs support outlet groups that can be managed as a logical unit, and user-level access control that restricts which administrators can control which outlets — essential in shared colocation environments where multiple tenants share a rack.

PDUs are available in single-phase and three-phase configurations. In Australia, single-phase power is 230V at 50Hz, delivered on a single active conductor. Single-phase PDUs are the most common type and are suitable for small to medium racks drawing up to approximately 7.3 kW (32A single-phase circuit). Most server rooms and edge deployments use single-phase PDUs because the building's electrical supply is single-phase and the per-rack power density does not justify three-phase infrastructure.

Three-phase PDUs are used in data centres and high-density environments where per-rack loads exceed what a single-phase circuit can deliver. Three-phase power in Australia provides 415V between phases and 230V between any phase and neutral. A three-phase 32A PDU can deliver approximately 22 kW — three times the capacity of a single-phase 32A circuit. Three-phase PDUs distribute the load across all three phases and include per-phase metering to ensure balanced loading. Unbalanced three-phase loads waste capacity and can cause neutral overload in some configurations, so the ability to monitor and balance phase loading is a key advantage of metered and switched three-phase PDUs.

PDU outlet types must match the power inputs of the equipment being powered. The most common outlet types in rack PDUs are IEC 60320 C13 (10A, used by servers, switches, and most IT equipment) and IEC 60320 C19 (16A, used by high-draw devices like blade chassis, large UPS units, and high-performance servers with redundant PSUs). A typical rack PDU might offer 20-24 C13 outlets and 4-6 C19 outlets to accommodate a mix of equipment. In Australian deployments, you may also encounter PDUs with AS/NZS 3112 (Type I) outlets — the standard Australian three-pin socket. These are useful for powering equipment with Australian plugs such as desktop computers, monitors, and consumer-grade networking equipment, but they are less common in professional data centre racks where IEC connectors are the standard.

Correctly sizing a PDU requires understanding three key parameters: the total power draw of the rack, the circuit capacity available, and the desired redundancy model. Start by adding up the maximum power consumption (in watts) of every device in the rack — use the nameplate ratings or, better yet, the actual measured draw if upgrading an existing deployment. Convert watts to amps using the formula: Amps = Watts ÷ Volts (so 2,300W ÷ 230V = 10A in Australia). Then apply a derating factor — best practice is to load a PDU to no more than 80 per cent of its rated capacity to allow for power spikes and future growth. A 32A PDU should therefore carry no more than 25.6A sustained load, equating to approximately 5,900W.

For redundancy, most critical environments use an A+B power feed model with two PDUs per rack, each on a separate circuit from a separate UPS. Each server is connected to both PDUs via its dual power supplies. In this configuration, each PDU must be capable of carrying the full rack load independently — because if one PDU or circuit fails, the other must sustain all equipment without overloading. This means you effectively need double the PDU capacity of your actual load. When specifying dual-PDU configurations, ensure the PDU form factor fits your rack — most racks accommodate two vertical 0U PDUs mounted on the rear uprights, preserving all rack units for equipment.