APC by Schneider Electric Rack PDU 2G Switched and Metered ZeroU AP8953

The APC Rack PDU 2G Switched and Metered ZeroU AP8953 is a premium power distribution unit engineered for ultra-high-density compute environments where per-outlet visibility and granular control are mission-critical. Designed with zero vertical footprint (ZeroU architecture), this three-phase PDU mounts horizontally within standard 19-inch racks without consuming vertical rail space, maximizing usable rack density. The unit delivers up to 11kW of managed power capacity with simultaneous metering and switching at the individual outlet level, enabling real-time energy consumption tracking and remote outlet-level power cycling for each connected device.

Built for the modern AI and GPU compute era, the AP8953 addresses the critical power management challenge facing enterprises deploying high-density GPU clusters and machine learning accelerator arrays. Each outlet is independently monitored and controllable, allowing data center operators to identify power anomalies, implement granular load balancing, and prevent cascading power events through intelligent outlet-level sequencing. The metered capability provides detailed energy analytics per GPU or accelerator card, supporting chargeback models and capacity planning in shared infrastructure environments.

The PDU integrates advanced networking and management features including Intelligent Power Manager (IPM) compatibility, SNMP/web interface controls, and environmental sensor support. Dual-input three-phase connectivity ensures resilience, while the compact form factor preserves critical rack real estate in space-constrained high-performance computing facilities.

Ideal for

• AI/ML training clusters requiring per-GPU power monitoring and load balancing across dual-GPU or quad-GPU node designs
• High-frequency trading and financial compute environments needing microsecond-level power event detection and individual server isolation
• Cryptocurrency mining and blockchain validation operations with per-ASIC outlet metering and remote power cycling for fault recovery
• Dense HPC supercomputing clusters where per-accelerator energy accounting enables workload optimization and thermal management
• Multi-tenant cloud data centers implementing granular billing models based on actual per-device power consumption
• Research computing facilities deploying heterogeneous GPU accelerators requiring real-time outlet-level diagnostics and controlled power sequencing

Technical specifications

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