What happened
Schneider Electric announced the RapidScale Modular Data Center Reference Design 115, a validated blueprint for high-density AI infrastructure that integrates the NVIDIA Vera Rubin platform with advanced liquid cooling. The company reported that this prefabricated approach reduces onsite installation time by 80% compared to traditional builds, aiming to streamline the deployment of AI factories globally.
Schneider Electric has unveiled RapidScale Modular Data Center Reference Design 115, a pre-engineered blueprint specifically designed for high-density AI infrastructure. According to Control Global, the design pairs the NVIDIA Vera Rubin platform with liquid cooling technology intended to maximize processing power while maintaining energy efficiency. The system integrates power, cooling, IT, software, and lifecycle services into a scalable, industrialized delivery model.
A key component of this announcement is a co-authored research paper with hyperscale developer Compass Datacenters, titled 'Prefabricated White Space Module Time Study: Lessons from a High-Density Liquid-Cooled Deployment.' The study, conducted at a Compass hyperscale data center campus, found that prefabricated modular pods reduced onsite installation time from 75.5 hours to 15 hours per pod. This represents an 80% reduction in labor time, achieved by moving critical assembly, integration, and verification into a controlled factory environment rather than relying on onsite coordination of multiple trades.
Schneider Electric states that this modular approach is a foundational step toward supporting the NVIDIA DSX POD reference design. This integration is designed to allow for factory-level validation of Vera Rubin AI infrastructure before it is shipped to the site, which the company claims will accelerate the time to first for new AI deployments. The reference design is currently validated to ANSI standards for U.S. deployments, with plans to extend support to IEC standards for global implementations in the future.
The company highlighted its supply chain capabilities as a critical factor in reducing project risk. Schneider Electric noted that its manufacturing capacity has grown fourfold since 2025, now covering eight primary sites totaling 1.35 million square feet, along with three additional facilities and 12 partner sites across 10 countries. Approximately 90% of the equipment used in RapidScale is manufactured in-house, which the company says ensures predictable performance and supply chain certainty.
Source details: controlglobal.com
Why it matters
This development addresses the critical bottleneck of skilled labor shortages and complex system integration in AI data center construction. By shifting assembly to factory-controlled environments, operators can deploy capacity faster and more predictably. This is significant as AI workloads are projected to drive a 170 GW increase in global data center capacity by 2030, making standardized, repeatable infrastructure essential for meeting demand without compromising on power efficiency or regulatory compliance.
The rapid expansion of AI infrastructure is straining traditional data center construction methods. Schneider Electric estimates that global data center capacity will grow by 170 GW by 2030, reaching a total of 300 GW, which is 2.3 times the 2025 installed base. With AI workloads expected to account for approximately 65% of this demand, the industry faces significant pressure to deploy high-density computing efficiently.
Traditional 'stick-built' data centers are increasingly difficult to scale due to skilled labor shortages and the complexity of integrating high-power, liquid-cooled systems. The 80% reduction in onsite installation time reported by Schneider Electric and Compass Datacenters offers a practical solution to these workforce bottlenecks. By standardizing the 'white space' into prefabricated modules, operators can reduce onsite complexity and accelerate the time it takes to bring AI capacity online.
This move also has implications for energy management and regulatory compliance. As AI factories draw more power from the grid, the need for precise thermal management and energy efficiency becomes paramount. The RapidScale design’s focus on smart controls and fluid cooling aims to deliver peak performance without overheating or wasting power, which is critical for meeting sustainability goals and grid capacity constraints.
Interactive Mechanism: How It Actually Works
Explore the underlying technology behind this development interactively.
What should a useful AI forecast state?
What to watch next
Monitor the extension of this reference design to support IEC standards for global deployments beyond the U.S. Also watch for further integration with the NVIDIA DSX POD reference design, which aims to enable factory validation of AI infrastructure before shipment, potentially further reducing the time to first for new AI factories.
The transition from U.S.-specific ANSI validation to global IEC standard support will be a key indicator of the design's international adoption. As AI infrastructure expands globally, compliance with local electrical and safety standards is a major hurdle for prefabricated solutions.
Further details on the integration with the NVIDIA DSX POD reference design will be important to understand how factory validation impacts the overall deployment timeline. If this process significantly reduces the time between hardware shipment and operational AI workloads, it could become a standard requirement for large-scale AI factory projects.
The scalability of Schneider Electric’s manufacturing capacity will be tested as demand for AI infrastructure continues to rise. The company’s claim of fourfold growth in manufacturing capacity since 2025 will need to be sustained to meet the projected 170 GW increase in data center capacity by 2030.