Sharon AI Reborns Former Paper Mill Into Major AI Data Hub

Dominic Jainy is a leading authority in data center infrastructure and technology real estate, specializing in the revitalization of industrial sites into high-performance computing hubs. With years of experience overseeing massive cloud expansions, he provides a unique perspective on Sharon AI’s recent AU$50 million acquisition of the former Ettamogah paper mill. Today, we discuss the strategic shift toward regional New South Wales, the deployment of 64,000 Nvidia GPUs, and the technical hurdles of bringing 212MW of capacity to an underutilized brownfield site.

The former paper mill in Ettamogah is transitioning into a high-tech center for AI cloud computing. What makes this specific brownfield site such a strategic asset for Sharon AI’s expansion in the Australian market?

This AU$50 million acquisition is a masterclass in repurposing legacy infrastructure for the modern AI era. The site, which was originally built in the 1980s, offers a robust power and logistics footprint that would be incredibly time-consuming and expensive to develop from scratch in a greenfield project. By revitalizing an underutilized site that Visy previously utilized for logistics, Sharon AI effectively bypasses many of the initial construction hurdles. This move places them strategically in the southern New South Wales corridor, bridging the gap between their established operations in Sydney and Melbourne.

Local representatives have stressed the need to “get the rules right” to protect the community as these facilities scale. How do you address these regulatory concerns while transforming a historic industrial landmark?

Balancing high-tech growth with community protection is essential when transitioning a historic site like the former Norske Skog mill. The concerns voiced by local MP Justin Clancy highlight the need for transparent integration into the local power grid and environment to ensure long-term sustainability. Since Visy will remain on-site as a tenant for warehouse and logistics space, the transition functions more as a strategic collaboration rather than a total industrial displacement. This hybrid model ensures that the 212MW capacity expansion enhances the region’s economic standing while respecting the expectations of the Ettamogah community.

Sharon AI plans to deploy 64,000 Nvidia GPUs by mid-2027. What are the primary technical and operational challenges of implementing this level of computing power in a regional location?

Deploying 64,000 Nvidia GPUs by mid-2027 is an enormous technical feat that requires meticulous planning regarding heat management and consistent energy uptime. Scaling to 212MW of capacity means the facility must be prepared for the most demanding generative AI workloads currently existing in the global market. Moving beyond their current partnerships with Equinix and CDC, this site represents Sharon AI’s transition toward self-managed, massive-scale infrastructure. It is a bold statement that positions regional New South Wales as a primary contender in the global cloud market rather than just a secondary capacity user.

What is your forecast for the Australian regional data center market?

I predict a significant surge in regional investments as major tech firms move away from saturated metropolitan zones to tap into the industrial power capacity of locations like Albury. By 2028, these regional hubs will likely become the backbone of Australia’s AI capabilities, offering a more sustainable and cost-effective alternative to traditional urban data centers. The success of the Ettamogah project will serve as a definitive blueprint for how legacy mills and factories across the country can be reborn as digital infrastructure leaders. This shift will ultimately provide regional communities with a new identity as essential players in the global technology race.

Explore more

6 Essential MacOS Settings to Change on Your New Mac

Customizing the delay before a key begins to repeat allows power users to delete text or input characters with greater precision and speed. Unboxing a new Mac is a ritual for many professionals in 2026, representing a fresh start for productivity and creative endeavors. However, the default configurations provided by Apple are often designed for a broad, general audience, which

Ethereum Leads the Way in $1.19 Billion Crypto Liquidation Event

The lopsided nature of the crypto derivatives market became clear when data showed that long liquidations outpaced short liquidations by a staggering eight-to-one ratio. This massive deleveraging event, which occurred between late October 8 and the early hours of October 9, 2026, resulted in approximately $1.19 billion in leveraged positions being wiped off the books in what traders commonly refer

macOS Tahoe vs. Sequoia: Which Version Should You Choose?

The transition from Sequoia to Tahoe includes a redesigned Control Center that supports multiple pages and deeper integration with third-party software controls. Since the arrival of macOS 27 Golden Gate on September 14, 2026, the landscape of Apple operating systems has shifted toward a future dominated by Apple Silicon, yet many users find themselves caught between the legacy stability of

Microsoft Launches Project Zenith to Debloat Windows 11

The decision to keep the Windows Dev Config script as a niche tool on GitHub creates a significant barrier for average users who desire a more professional operating system experience. Microsoft has recently initiated Project Zenith, an internal effort aimed at refining the Windows 11 experience by stripping away the “bloat” that has long frustrated its user base. This project

How Are AI-Driven Campaigns Reshaping Modern Cyber Threats?

Recent investigations into CyberXero reveal a dual-track strategy that combines mass e-commerce exploitation with targeted strikes on critical infrastructure. This approach represents a fundamental shift in how digital adversaries operate in 2026, moving away from labor-intensive manual methods toward a highly efficient, automated model. By leveraging sophisticated artificial intelligence orchestration layers, these threat actors can now execute high-volume attacks while