Raritan PX3-5565 Metered-by-Outlet High-Voltage PDU

The Raritan PX3-5565 is a 415V three-phase high-voltage power distribution unit designed to meet the demanding requirements of next-generation data center GPU server infrastructure. The unit delivers metered power management at individual outlet granularity, enabling real-time monitoring and intelligent load balancing across AI accelerator racks. This architecture addresses the critical infrastructure challenge of high-density GPU deployments where precise power visibility and control per compute node is mandatory for operational efficiency and thermal management.

Built for European and APAC high-density computing environments, the PX3-5565 provides comprehensive per-outlet metering of voltage, current, power factor, and energy consumption. The device integrates intelligent circuit protection, remote power management capabilities, and detailed outlet-level telemetry that feeds directly into data center infrastructure management (DCIM) platforms. This granular visibility enables administrators to optimize power allocation, identify power anomalies, and manage capacity constraints in real-time across GPU cluster deployments.

The unit supports both intelligent outlet control (individual socket on/off switching, outlet sequencing, and load shedding policies) and advanced monitoring features including harmonic analysis and reactive power measurement. Raritan's firmware and management integration ensure seamless interoperability with leading DCIM and intelligent power management ecosystems, critical for data centers where GPU workloads create complex, dynamic power consumption patterns.

Ideal for

• AI model training clusters requiring per-GPU power monitoring for cost allocation, workload optimization, and carbon footprint tracking across multi-tenant environments
• High-performance computing (HPC) GPU server racks where outlet-level circuit protection and intelligent load shedding prevent cascading power failures during peak compute phases
• Machine learning inference infrastructure deployments needing real-time power telemetry to correlate compute performance metrics with energy consumption per accelerator
• Data center capacity planning and power management for GPU-dense facilities where per-outlet metering enables predictive load balancing and prevents PDU overloads
• Research computing and generative AI facilities requiring detailed energy accountability and per-workload power consumption attribution for operational billing and efficiency audits
• Enterprise GPU server consolidation projects where outlet-level visibility drives infrastructure right-sizing decisions and thermal management optimization

Technical specifications

Available from Omnixon Global. Submit an RFQ and our team will confirm configuration and availability for your order.