SoftBank Targets Infrastructure Dominance with Proposed $100 Billion Data Center Spinoff
SoftBank Group Corp. is reportedly preparing to launch an autonomous, AI-driven data center construction firm named Roze. If the reports from the Wall Street Journal and the Financial Times hold true, SoftBank aims to take the venture public on a U.S. exchange by the end of this year, targeting a staggering $100 billion valuation. This move represents a strategic pivot for CEO Masayoshi Son, shifting the conglomerate’s focus from passive venture capital to the direct industrialization of AI-critical infrastructure.
The Convergence of Robotics and Infrastructure
The core innovation behind Roze lies in the integration of robotic automation into the construction lifecycle. Building massive data centers remains a labor-intensive, multi-year endeavor prone to delays and supply chain bottlenecks. By leveraging robotics—likely sourced from SoftBank’s existing acquisitions such as the precision welding arms gained from its $5.4 billion purchase of ABB Group’s robotics division—Roze could theoretically compress construction timelines and drastically lower overhead costs.
Industry analysts suspect that Roze will serve as a hub for SoftBank’s existing infrastructure assets, most notably SB Energy. The unit carries significant pedigree, having already secured contracts for massive high-capacity power centers, such as the 1.2-gigawatt Stargate project for OpenAI. Integrating energy generation with construction capability allows SoftBank to offer a vertically integrated turnkey solution for the hyperscalers responsible for the current AI gold rush.
Solving the Cable Bottleneck with Hardware Innovation
The vision extends beyond site assembly; SoftBank is actively addressing the physical limitations of server maintenance. Robots are notoriously clumsy when navigating the chaotic rat’s nest of fiber-optic and power cables inherent in traditional server racks. To bypass this, SoftBank has pioneered a proprietary cableless server rack design, which utilizes conductive metal bars and simplified network connectors instead of fragile, complex wiring.
Testing for this hardware is already underway in Hokkaido, Japan. If successful, this technology could standardize a new class of robot-ready data centers, where autonomous systems handle everything from server rack installation to predictive maintenance and component replacement. This shift toward self-maintaining facilities is the logical next step in mitigating the massive operational costs associated with the modern AI cloud.
Strategic Implications and Financial Ambitions
The $100 billion valuation target, while ambitious, reflects the premium the market currently places on AI-supporting infrastructure. SoftBank is clearly positioning itself to be more than just a financier; it aims to be the architect of the physical world that sustains the digital intelligence revolution.
However, observers remain cautious about the timeline. Executives within the company have reportedly characterized the end-of-year IPO goal as aggressive, noting that the complexities of spinning off an industrial entity of this scale could lead to delays.
Furthermore, the capital raised from a successful Roze listing would likely serve as a massive war chest for further aggressive expansion. SoftBank has already signaled its intent to entrench itself in the industry through recent substantial investments in OpenAI and the acquisition of DigitalBridge Group. By controlling the power, the construction, the hardware, and the funding, SoftBank is attempting to build a self-sustaining ecosystem that dictates the speed at which the global AI infrastructure is deployed.