Nvidia’s big reveal: an AI data centre in a box, rapid set-up

By the Editorial Team · Friday, October 23, 2026

Nvidia has launched a modular AI data centre solution designed to speed up deployment and cut energy and water use, supported by Firmus's innovative cooling technology. These advancements address critical infrastructure challenges, enabling faster, more sustainable AI operations that could significantly boost economic output by reducing costly construction delays.

Nvidia has introduced a modular AI data centre solution aimed at accelerating deployment times and reducing resource consumption in artificial intelligence operations. The innovation was unveiled alongside related developments from Firmus, a company with close ties to Nvidia, including equity investment and supply agreements for AI chips.

Firmus, led by Curtis, is collaborating with the Australian Securities Exchange-listed Maas Group through its subsidiary JLE to deliver “Power Cubes”—modular units designed for rapid assembly in AI factory projects. The company highlights its proprietary cooling technology, which reportedly achieves 33 percent lower energy use and reduces water consumption by 99 percent when maintaining optimal temperatures for high-performance graphics processing units (GPUs).

This technology aims to address one of the critical bottlenecks in AI infrastructure: the cooling of GPUs, which are essential for AI computations but generate significant heat and typically require substantial power and water resources.

At a separate event in Berlin, Nvidia showcased its DSX AI factory platform, emphasizing efficiency improvements that include minimized water usage and reduced power requirements relative to the AI output delivered. One key component of the announcement was the DSX Pod, a modular system intended to simplify and speed up data centre construction and AI processing times.

Industry experts note that the speed of data centre development can translate directly into economic impact, citing that a one-gigawatt data centre infrastructure can generate tens of billions of dollars in annual revenue. Delays as short as three months may result in billions in lost income, underscoring the importance of Nvidia’s efforts to streamline AI data centre deployment.

These advancements come amid growing demand for AI capabilities and the corresponding need for scalable, sustainable infrastructure to support expanding computational requirements. By improving the efficiency and setup speed of AI data centres, Nvidia and its partners seek to maintain competitive advantages in the rapidly evolving AI sector.

This article was reported by synthesizing coverage from multiple independent news outlets to present a balanced, unbiased account. Read about our editorial standards. Spotted an error? Let us know.