How Energy Monitoring and Smart PDUs Contribute to More Efficient Data Centres
Discover how smart PDUs and inline meters improve rack-level energy monitoring, power management and efficiency in modern AI, HPC and data centre environments.
HPC & AISchleifenbauer PDU
Data centres are consuming more power than ever. AI infrastructure, high-performance computing (HPC), GPU servers and high-density racks are increasing power requirements, while operators need greater visibility into exactly where that energy is being used.
The challenge is no longer simply distributing enough power. Data centre operators need to measure, monitor and manage power at rack level, detect capacity issues before they become problems and integrate energy data into DCIM, BMS and other management platforms.
This is where intelligent Power Distribution Units (PDUs) and inline energy meters play an important role.
Smart PDUs provide detailed information about power consumption directly inside the rack. Inline meters can add similar monitoring capabilities to existing infrastructure without requiring the complete power distribution setup to be replaced.
In this guide, we explain how both technologies work, when to use them and how they can support more efficient data centre energy management.
What Is a PDU in a Data Centre?
A Power Distribution Unit (PDU) distributes electrical power to servers, switches, storage systems, GPU servers and other IT equipment inside a data centre rack.
A basic PDU primarily distributes power. An intelligent or smart PDU goes further by measuring electrical parameters and, depending on its configuration, providing remote monitoring and switching capabilities.
This gives data centre operators visibility into what is happening between the facility power infrastructure and the IT equipment consuming that power.
What Types of Data Centre PDUs Are Available?
PDUs can generally be divided into several categories:
Basic PDU: distributes power without remote energy monitoring.
Metered PDU: measures electrical consumption at input or branch level.
Monitored PDU: provides more granular monitoring, potentially down to individual outlets.
Switched PDU: allows outlets to be switched remotely.
Managed PDU: combines detailed monitoring, remote outlet control and network-based management.
The exact functionality depends on the PDU configuration.
Schleifenbauer PDU 5.0 systems, for example, can provide measurements at input, branch and outlet level, depending on the selected configuration. They can also integrate with external management environments using protocols such as REST API, MODBUS/TCP and SNMP.
Why Are Smart PDUs Important in Modern Data Centres?
A smart PDU turns power distribution into a measurable part of the data centre infrastructure.
Instead of only knowing how much electricity enters a facility or distribution board, operators can see how power is being consumed much closer to the actual IT load.
This can help data centre teams to:
monitor rack-level power consumption;
identify approaching capacity limits;
balance loads across phases and circuits;
analyse historical energy consumption;
configure alarms for abnormal conditions;
allocate energy consumption to racks or equipment;
remotely control selected outlets;
support capacity planning;
integrate power data into DCIM or BMS platforms.
These capabilities become particularly important as rack power densities increase.
AI and HPC infrastructure can create significantly higher and more dynamic loads than traditional enterprise IT equipment. Accurate rack-level monitoring therefore becomes an important part of capacity and energy management.
Smart PDUs sharing energy management data
What Is an Inline Energy Meter?
An inline energy meter is a device installed between the power source and the equipment or PDU being monitored.
Instead of replacing an existing PDU, an inline meter adds energy monitoring to the existing power infrastructure.
This makes inline meters particularly useful for data centre retrofit projects.
For example, a data centre may already have hundreds of basic PDUs installed. Replacing every PDU purely to gain monitoring functionality can be costly and operationally disruptive.
An inline meter can provide another option: keep the existing power distribution equipment and add measurement upstream.
Schleifenbauer’s Inline Meter 5.0 is designed for this type of retrofit application and can be installed between the power supply and an existing PDU.
What Does an Inline Meter Measure?
Depending on the configuration, an inline meter can provide measurements such as:
voltage (V);
current (A);
active power (kW);
apparent power (kVA);
energy consumption (kWh);
power factor;
frequency.
This data can then be made available to monitoring or management software for analysis, reporting and alarms.
Smart PDU vs Inline Meter: What Is the Difference?
Both solutions provide visibility into energy consumption, but they solve different infrastructure challenges.
Smart PDU
Inline Meter
Distributes and monitors power
Primarily adds measurement capability
Installed directly in the rack
Installed between the power source and existing equipment/PDU
Can provide input, branch or outlet monitoring
Useful for adding monitoring to existing infrastructure
Can include remote outlet switching
Typically used primarily for measurement
Suitable for new installations and upgrades
Particularly suitable for retrofit projects
Can replace an existing basic PDU
Can allow an existing PDU to remain in service
The right solution therefore depends on the data centre.
For a new rack or high-density deployment, an intelligent PDU is often the logical choice.
For an existing data centre with functioning basic PDUs, an inline meter can be an efficient way to introduce energy monitoring without redesigning the entire rack power infrastructure.
Inline Meter 5.0, retrofit Basic and legacy intelligent PDUs to intelligent
What Is the Difference Between a Metered and Switched PDU?
The main difference is control.
A metered PDU measures electrical parameters and energy consumption but does not necessarily provide remote outlet switching.
A switched PDU allows authorised users or management systems to remotely turn individual outlets on or off, depending on its configuration.
Remote switching can be useful for tasks such as:
remotely rebooting equipment;
controlling unused IT loads;
performing staged power-up sequences;
reducing inrush current;
managing equipment at unmanned or remote sites.
Schleifenbauer switched PDUs use bistable relays, which only require energy when changing state. They can also use Near Zero Voltage Switching and switching delays to support controlled switching and reduce the impact of inrush currents.
How Do You Measure Power Consumption Per Server Rack?
There are three common approaches to rack-level energy monitoring.
1. Intelligent or Monitored PDUs
A smart PDU can measure energy directly inside the rack.
Depending on the configuration, measurements can be available at:
Input level: total consumption entering the PDU.
Branch level: consumption of individual electrical branches.
Outlet level: consumption associated with individual PDU outlets.
Outlet-level measurement provides the highest granularity and can be particularly useful when energy consumption needs to be allocated to specific equipment or customers.
2. Inline Energy Meters
An inline meter can be installed upstream of an existing PDU.
This is particularly useful when the current rack PDUs do not provide the required monitoring capabilities but are otherwise still suitable for operation.
3. External Energy Meters
Energy meters can also be installed at distribution-board level.
This provides useful information about larger electrical circuits, although it generally offers less rack-level granularity.
In practice, many data centres use a combination of these approaches.
How Can PDU Data Be Integrated With DCIM and BMS Platforms?
Energy measurements become significantly more useful when they can be integrated into the data centre’s existing management environment.
Rather than creating another isolated monitoring platform, open communication protocols allow PDU and energy data to become part of a wider DCIM, BMS, NMS or energy management strategy.
Schleifenbauer’s current PDU and Inline Meter platform supports protocols including:
HTTP/HTTPS;
REST API;
MODBUS/TCP;
SNMP, including SNMPv3;
SNMP traps;
IPv4/IPv6;
Syslog;
CLI.
The exact interfaces available depend on the product and communication module.
This allows organisations to use rack-level measurements for dashboards, alarms, automated reporting, capacity planning and integration with other operational systems.
What Is EnerTree?
EnerTree is Schleifenbauer’s Data Centre Energy Management (DCEM) software for monitoring and managing PDU and energy data.
It can collect and visualise measurements such as energy consumption, voltage, current, temperature and rack load.
Depending on the architecture, EnerTree can also support:
historical energy analysis;
dashboards;
alarm thresholds;
reporting;
PUE analysis;
capacity planning;
device discovery;
integration with DCIM environments.
EnerTree is available as EnerTree Lite, embedded with the Controller Module, and EnerTree Platform for larger centralised environments. Schleifenbauer states that the platform architecture can scale to thousands of connected PDUs.
EnerTree DCEM Software
How Can Smart PDUs Help Improve Data Centre PUE?
Power Usage Effectiveness (PUE) compares the total energy consumed by a data centre with the energy consumed by its IT equipment.
The formula is:
PUE = Total Facility Energy / IT Equipment Energy
A theoretical PUE of 1.0 would mean that all facility energy is used directly by IT equipment.
A smart PDU does not automatically lower PUE. What it provides is the measurement data needed to identify opportunities for optimisation.
For example, rack-level monitoring can help operators:
identify underutilised equipment;
understand actual IT loads;
improve capacity planning;
detect abnormal consumption patterns;
balance electrical loads;
compare racks and zones;
analyse energy trends over time.
Combining accurate rack-level measurements with facility-level energy data creates a much clearer picture of where energy is being consumed.
That makes PDU data valuable not only for power management, but also for broader data centre energy-efficiency strategies.
Why Is Power Monitoring Important for AI and HPC Data Centres?
AI is changing the power profile of the server rack.
GPU-based systems can require significantly more power than conventional enterprise servers, making power capacity, monitoring and cooling increasingly interconnected.
A good example is the NVIDIA DGX H100.
NVIDIA specifies a maximum system power consumption of 10.2 kW for one DGX H100 system. The system occupies 8U and contains six 3.3 kW power supplies.
NVIDIA’s DGX SuperPOD design guidance describes configurations with up to four DGX H100 systems per rack. Four systems represent up to 40.8 kW of server power per rack before other rack equipment is considered.
At these densities, power distribution is no longer simply about providing enough sockets.
Operators need to consider:
available circuit capacity;
phase balancing;
redundancy;
connector types;
cable ratings;
continuous load;
peak demand;
cooling capacity;
real-time monitoring;
alarm thresholds;
future capacity.
NVIDIA’s own DGX H100 data centre guidance describes 415 VAC, 32 A, three-phase N+1 as the preferred power arrangement for high-density DGX H100 deployment patterns, although other configurations are possible depending on rack density and infrastructure.
This is exactly where high-density intelligent PDUs become important.
What Should You Look for in a PDU for High-Density and AI Racks?
There is no single PDU specification that is correct for every AI rack.
The PDU should be selected based on the actual server configuration, redundancy strategy, available power and electrical infrastructure.
For high-density environments, consider:
single-phase versus three-phase power;
maximum current and power capacity;
number and type of outlets;
outlet-level measurement;
remote switching requirements;
measurement accuracy;
branch monitoring;
redundancy architecture;
temperature monitoring;
network connectivity;
DCIM/BMS integration;
cybersecurity requirements;
future rack power requirements.
The goal should be to select a PDU based on the actual electrical design of the rack, rather than selecting one purely on its maximum current rating.
PDU Monitoring and Cybersecurity
As PDUs become intelligent network devices, cybersecurity becomes part of power infrastructure management.
A network-connected PDU can expose measurement data, configuration information and, in switched configurations, remote switching functions.
Access therefore needs to be properly secured.
Schleifenbauer PDU 5.0 and EnerTree support security and management functionality including HTTPS, TLS 1.3, user and role management, LDAP/Active Directory, SNMPv3 and SSH.
For data centre operators, intelligent PDU security should therefore be considered alongside the broader network and infrastructure security architecture.
When Should You Choose an Inline Meter Instead of Replacing a PDU?
An inline meter is particularly useful when:
existing PDUs are still operational;
energy monitoring is currently unavailable;
replacing PDUs would cause unnecessary disruption;
rack-level consumption needs to be measured;
a phased data centre upgrade is preferred;
existing equipment needs to be integrated into a modern monitoring platform.
This makes inline energy monitoring particularly relevant for brownfield and retrofit data centre projects.
Instead of replacing a large installed base at once, operators can introduce energy visibility incrementally.
From Power Distribution to Data-Driven Energy Management
Modern data centre power infrastructure is becoming increasingly intelligent.
A PDU is no longer simply a row of electrical outlets inside a server rack. Intelligent PDUs can become a detailed measurement point between facility power and IT infrastructure.
Inline meters provide another route to the same objective: better visibility into where energy is being consumed.
For traditional enterprise environments, this supports capacity planning and energy management. For colocation providers, granular measurement can support reporting and allocation. For edge locations, remote monitoring reduces the need for physical intervention.
And for AI and HPC environments, where rack densities can reach tens of kilowatts, accurate real-time energy monitoring is becoming increasingly important.
Schleifenbauer develops configurable intelligent PDUs, Inline Meters and EnerTree DCEM software for these environments, with monitoring from input to outlet level and integration with existing data centre management platforms.
Schleifenbauer – Infographic PDU 5.0 EnerTree Platform vs EnerTree lite
Frequently Asked Questions About PDUs and Inline Energy Monitoring
What is a PDU in a data centre?
A Power Distribution Unit (PDU) distributes electrical power to servers and other IT equipment inside a rack. Intelligent PDUs can additionally measure energy consumption, monitor electrical parameters and, depending on the configuration, remotely switch outlets.
What is a smart PDU?
A smart PDU or intelligent PDU is a network-connected Power Distribution Unit that combines power distribution with monitoring, management or switching functionality. It can provide data such as voltage, current, power and energy consumption to data centre management systems.
What is the difference between a metered and monitored PDU?
A metered PDU provides electrical measurements at defined measurement points such as the input or branches. A monitored PDU provides more detailed remote monitoring and can, depending on its configuration, provide measurements down to individual outlet level.
What is the difference between a smart PDU and an intelligent PDU?
The terms smart PDU and intelligent PDU are generally used to describe PDUs with network-connected monitoring or management capabilities. Exact functionality varies by manufacturer and model.
What is an inline energy meter?
An inline energy meter is installed between an electrical power source and the equipment being monitored. In a data centre, it can be placed upstream of an existing rack PDU to add energy monitoring without necessarily replacing the PDU.
Can I add energy monitoring to an existing basic PDU?
Yes. An inline energy meter can be used to add energy measurement to existing power distribution infrastructure. This can be particularly useful when retrofitting existing data centre racks.
How do you measure the energy consumption of a server rack?
Rack energy consumption can be measured using an intelligent PDU, an inline energy meter or external metering at the electrical distribution level. Intelligent PDUs generally provide the most detailed rack-level information, especially when outlet-level monitoring is available.
Can a PDU measure individual servers?
A PDU with outlet-level monitoring can measure the consumption associated with individual outlets. If one server or power supply is connected to a dedicated monitored outlet, this provides granular information about its electrical load.
Can smart PDUs integrate with DCIM software?
Yes. Intelligent PDUs can integrate with DCIM, BMS and monitoring platforms through protocols and APIs. Schleifenbauer PDU 5.0 supports interfaces including REST API, MODBUS/TCP and SNMP.
How much power does an NVIDIA DGX H100 consume?
NVIDIA specifies a maximum system power consumption of 10.2 kW for the DGX H100. NVIDIA’s data centre design documentation describes configurations of up to four DGX H100 systems per rack, representing up to 40.8 kW of server power.
Do AI servers require special PDUs?
Not necessarily a special category of PDU, but high-density AI infrastructure can require significantly higher rack power capacity. The PDU configuration therefore needs to match the rack’s voltage, current, redundancy, outlet, monitoring and power-density requirements.
Can smart PDUs help reduce data centre energy consumption?
Smart PDUs provide detailed energy data that can help operators identify inefficient equipment, unused capacity and abnormal consumption. The PDU itself does not automatically reduce energy consumption, but the data it provides can support optimisation decisions.
What is the best PDU for a data centre?
There is no single best PDU for every data centre. The correct configuration depends on rack power density, voltage, current, redundancy, outlet types, measurement requirements, remote switching, monitoring integrations and future capacity needs.
Smarter Energy Monitoring Starts at Rack Level
As rack densities continue to increase, understanding exactly where power is being consumed becomes essential.
Whether you are building a new AI-ready data centre, upgrading existing racks or adding energy monitoring to legacy infrastructure, the right combination of intelligent PDUs, inline energy meters and energy management software provides the data needed to make informed decisions.
Need help selecting the right PDU or Inline Meter configuration for your racks? Contact Schleifenbauer to discuss your power, monitoring and integration requirements.
To provide the best experiences, we use technologies like cookies to store and/or access device information. Consenting to these technologies will allow us to process data such as browsing behavior or unique IDs on this site. Not consenting or withdrawing consent, may adversely affect certain features and functions.
Functional
Always active
The technical storage or access is strictly necessary for the legitimate purpose of enabling the use of a specific service explicitly requested by the subscriber or user, or for the sole purpose of carrying out the transmission of a communication over an electronic communications network.
Preferences
The technical storage or access is necessary for the legitimate purpose of storing preferences that are not requested by the subscriber or user.
Statistics
The technical storage or access that is used exclusively for statistical purposes.The technical storage or access that is used exclusively for anonymous statistical purposes. Without a subpoena, voluntary compliance on the part of your Internet Service Provider, or additional records from a third party, information stored or retrieved for this purpose alone cannot usually be used to identify you.
Marketing
The technical storage or access is required to create user profiles to send advertising, or to track the user on a website or across several websites for similar marketing purposes.