Modular Beats Monolithic: How Rimac and Ecoblox Are Putting High-Density Battery Modules Inside the Data Center Shell

On October 8, 2026, Rimac Technology and ECOBLOX brought a significant idea into public view: high-density battery modules, power electronics, UPS functions, cooling, IT infrastructure, and monitoring can be engineered as one modular data center solution for AI workloads. The partnership combines Rimac’s UNI battery modules and high-voltage expertise with ECOBLOX’s prefabricated data center architecture and DCIM platform.

The timing is important. Grid interconnection queues are lengthening, transformer and switchgear lead times remain difficult to predict, and the power protection supply chain is under pressure from simultaneous demand across data centers, electrification, and utility-scale storage. Building a separate battery room sized for a facility’s theoretical three-year peak may create another long-lead construction package before the first GPU rack is even installed.

The more practical alternative is a modular power block that arrives factory-integrated, tested, and ready to connect to the data center shell. This does not eliminate the need for utility capacity, generators, UPS systems, or code compliance. It changes how those systems are deployed: in repeatable stages aligned with actual compute demand.

Why Now: Redundancy and Thermal Management Are Becoming One Design Problem

The traditional data center separates functions into dedicated spaces: a central UPS room, a battery room, mechanical plant, electrical room, and white space. That arrangement remains effective for many Tier III and Tier IV facilities. It also creates long cable runs, multiple field interfaces, separate commissioning activities, and a large amount of infrastructure that may sit underused for years.

AI changes the economics. A conventional enterprise rack may draw 5–15 kW. AI racks can reach 30, 50, or even 100 kW, depending on the GPU platform, power shelves, networking equipment, and workload. At those densities, a battery system is not merely a backup accessory. It is part of the facility’s power-conversion and thermal-management architecture.

Redundancy must also be considered at the module and block level. A monolithic 2 MW battery room can provide substantial capacity, but a failure in shared switchgear, controls, cooling, or distribution may affect a large portion of the facility. A modular architecture can reduce that blast radius by dividing capacity into independently isolated blocks. The precise arrangement still depends on the required fault-tolerance model: N+1 may be sufficient for one design, while 2N or distributed A/B paths may be required for another.

Thermal management adds another constraint. Batteries produce heat during charging, discharging, and conversion. Liquid-cooled GPU systems reject a much larger heat load into coolant distribution units, heat exchangers, and facility-water loops. Integrating battery modules into the same shell makes coordination easier, but it does not mean the battery cooling loop and IT liquid-cooling loop should automatically be combined.

The design must define heat rejection capacity, operating temperature limits, leak detection, isolation valves, coolant compatibility, and failure behavior. A battery block should remain safe and serviceable if a GPU cooling loop is offline. Likewise, a coolant leak in the IT area should not disable the power-protection system without a controlled isolation sequence.

What Pre-Integration Actually Buys

Pre-integration is more than placing several products in the same container. It means the battery modules, BMS, power conversion, UPS interface, cooling equipment, controls, and monitoring points are designed to work together before they arrive onsite.

That creates five practical advantages:

1. Factory-tested interfaces

The highest-risk commissioning problems often occur at the boundaries between systems. A factory-integrated module can test communication between the BMS, UPS, DCIM platform, thermal controls, alarms, and emergency shutdown circuits before shipment.

Rimac and ECOBLOX specifically describe streaming battery telemetry into a DCIM environment. That matters because operators need more than a battery percentage. They need state of charge, state of health, temperature, alarm status, isolation status, charge/discharge limits, and service history in one operational view.

2. Known thermal behavior

A separate battery room requires field validation of ventilation, cooling, temperature gradients, and clearance. A pre-engineered block can be tested against a defined operating envelope.

The vendor should still provide heat-load data at standby, charging, maximum discharge, and end-of-life conditions. For a high-density site, “battery capacity” without corresponding heat-rejection data is incomplete engineering.

3. Fewer field electrical connections

Every field connection is a potential installation error, inspection delay, or future maintenance issue. Factory-built battery cabinets and power blocks can reduce onsite terminations and provide documented torque values, labeling, protection settings, and test records.

This does not remove the need for qualified electrical installation. It reduces the number of variables that must be resolved in the field.

4. Shorter commissioning

Factory acceptance testing, or FAT, can validate protection trips, communications, simulated battery faults, transfer sequences, emergency stops, and thermal alarms before the equipment reaches the site. Site acceptance testing, or SAT, then focuses on utility connection, generator interaction, grounding, fire systems, network integration, and load-bank performance.

That sequence can compress the critical path compared with assembling a battery room, UPS system, cooling package, and monitoring platform from separate vendors.

5. More predictable lead times

A modular block still depends on batteries, switchgear, inverters, transformers, and controls. It is not immune to supply-chain constraints. However, a repeatable design gives the integrator a better opportunity to reserve complete blocks, standardize components, and avoid redesigning the battery room for every phase.

Containerized BESS Is Not the Same as an Integrated Battery Module

A containerized BESS typically places batteries, a power-conversion system, thermal management, fire protection, and controls in an outdoor enclosure. It may support 1 MW to hundreds of megawatts across multiple units and can serve a campus, microgrid, peak-shaving program, or utility-interconnection strategy.

That architecture is valuable when the battery system needs to operate as a grid asset or remain physically separate from the data hall. It can also simplify future expansion if the site has adequate outdoor space and electrical capacity.

An integrated battery module inside the data center shell serves a different purpose. It is designed around the electrical, mechanical, and operational requirements of a specific compute block. The battery may provide ride-through, generator bridging, power smoothing, peak support, or a combination of functions. Its controls are coordinated with the data center’s UPS, cooling, IT load, and DCIM systems.

Neither approach is automatically superior. The decision depends on runtime, power quality requirements, site constraints, fire-code strategy, utility programs, and the desired relationship between backup power and grid services.

For scale, the design brief should state capacity in clear increments. A proposed block might be 0.25 MW, 0.5 MW, or 1 MW of usable discharge power, but those figures must come from the final equipment selection. Rimac and ECOBLOX have not publicly disclosed the power rating, energy capacity, runtime, or rack density of their joint demonstration system. Operators should not infer those specifications from the partnership announcement.

As a reference point, Vertiv’s EnergyCore battery cabinets publish model-level power ratings from 108 kW to 263 kW, with five- or seven-minute runtime configurations depending on the model. Vertiv also publishes UPS systems with double-conversion efficiency up to 97.5% and dynamic operating modes up to 99%. These figures demonstrate why the complete power path matters: battery capacity, UPS efficiency, rack load, and cooling overhead must be modeled together.

The Modular Deployment Roadmap

A facility manager evaluating integrated battery architecture should take five steps before approving a deployment.

1. Define the compute block before sizing the battery block.
   Document the initial IT load, expected rack density, growth rate, required runtime, and power quality objectives. A 50 kW rack supporting AI training has a different ride-through profile from a 10 kW enterprise rack. Separate continuous load, transient load, startup load, and future expansion capacity.

2. Set a modular capacity schedule.
   Instead of building a battery room for a speculative three-year peak, define Phase 1 and future blocks. For example, a site may deploy 0.5 MW of protected compute capacity today and reserve space, busway, cooling connections, and switchgear provisions for a second and third block. The expansion path must be physical and electrical, not merely a line on a planning document.

3. Demand assembly-level safety evidence.
   Verify the complete system, not only the cells or battery modules. UL 9540A is a test method for evaluating thermal-runaway fire propagation at cell, module, unit, and installation levels. Ask for evidence applicable to the exact chemistry, enclosure, cooling method, module arrangement, and installation conditions. Also verify the appropriate UL 9540 system certification, UL 1973 battery-system listing, NFPA 855 requirements, and the authority having jurisdiction’s review criteria.

4. Validate liquid cooling and battery thermal boundaries.
   Require heat-load curves, coolant specifications, leak detection, isolation procedures, ventilation assumptions, and emergency shutdown logic. Confirm that battery cooling remains available during IT cooling maintenance and that a battery fault cannot propagate through a shared coolant or control system.

5. Make FAT, SAT, and service access contractual requirements.
   FAT should include BMS-to-DCIM communications, UPS transfer behavior, alarm mapping, emergency stops, protection settings, thermal alarms, and simulated module isolation. At the site, SAT should test utility, generator, bypass, load-bank, grounding, fire detection, and A/B distribution behavior. Confirm that individual modules can be isolated and replaced without taking down the entire block.

A Practical Commercial Case for Phased Capacity

Consider an operator deploying an AI cluster in stages. The first customer commitment requires 300 kW of protected IT load. A conventional approach might construct a battery room and electrical distribution system for 1 MW because that is the projected requirement three years from now.

That approach creates stranded capital, unused cooling capacity, and a larger commissioning scope. It also exposes the operator to technology risk: battery chemistry, UPS topology, rack voltage, and cooling requirements may change before the third year.

A modular approach installs the first power block for the committed deployment while reserving space and interfaces for future blocks. Capital spending follows revenue-producing compute. The operator can evaluate actual load growth, runtime requirements, maintenance data, and utility conditions before ordering the next module.

This is the central benefit of modular power protection: it turns capacity planning into a controlled sequence of decisions rather than one oversized construction bet.

Modern Real-Time Solutions should provide visibility from the rack to the battery module, UPS, switchgear, cooling system, and utility interface. Ace Real Time Solutions supports that broader approach through power audits, UPS and battery design, professional installation, and ongoing monitoring. Review our resources on AI power swings, rack-level power architecture, and smart UPS monitoring.

Frequently Asked Questions

What is an integrated battery module in a modular data center?

An integrated battery module is a factory-engineered energy-storage and power-protection assembly installed within the same modular data center architecture as the UPS, power conversion, monitoring, cooling, and IT infrastructure. It differs from a separate battery room or outdoor BESS because its electrical, thermal, control, and service interfaces are designed as one repeatable compute block.

How does modular battery architecture improve data center scalability?

Modular architecture allows operators to add capacity in phases tied to actual compute deployment. Instead of installing a battery room for a future peak load, the facility can deploy an initial block and add a second or third block when rack demand, customer commitments, and utility capacity justify the investment. This reduces stranded capital while preserving a defined expansion path.

What should be verified before deploying integrated battery modules?

Verify the exact module and assembly configuration, usable MW and MWh ratings, runtime at the intended rack load, UPS efficiency, BMS and DCIM compatibility, liquid-cooling behavior, maintenance clearances, module-level isolation, FAT and SAT procedures, and assembly-level safety evidence. Confirm UL 9540, UL 1973, UL 9540A test applicability, NFPA 855 requirements, and Tier III or Tier IV design objectives with the project engineer and AHJ.

Design the Next Power Block Around Your Actual Load

AI infrastructure is scaling faster than many construction programs can respond. A separate, oversized battery room may still be the right choice for a large campus, but it should not be the default.

Pre-integrated modular power protection offers a more disciplined path: factory-tested interfaces, predictable thermal behavior, fewer field connections, shorter commissioning, and capacity that can grow with deployed compute. The architecture must be engineered carefully, but the commercial logic is direct: build what the facility needs now, and make the next block easier to add.

Visit acerts.com to review power protection solutions, download technical information, or request an enterprise solution design. Ace Real Time Solutions can help assess your rack density, runtime target, redundancy model, cooling constraints, and expansion plan through a comprehensive power audit.