Wardan Core: A Cold, Deep Future for Sustainable AI Infrastructure
By Edward Shepherd - Founder, Wardan Core Initiative
Australia’s AI future is coming, fast, hot, and thirsty.
As artificial intelligence accelerates, so too does its demand for energy and water. The cooling needs of data centres are already driving infrastructure decisions, emissions budgets, and grid strain globally. Australia, uniquely positioned between vast ocean shelves and an emergent offshore renewables sector, is being handed an opportunity on a silver platter: to rethink the foundation of its digital infrastructure from the seabed up.
🌊 Introducing Wardan Core
Wardan Core is a conceptual initiative proposing modular, submerged AI data infrastructure cooled by deep cold seawater and powered by offshore renewable energy, ideally wind, but potentially wave or floating solar where viable. Its name, drawn from the Noongar word for “ocean,” anchors it in the unique identity of our southern coastlines, while pointing toward a future that is globally exportable.
This is not fantasy. Microsoft has already trialled a version of this with Project Natick off the coast of Scotland. In China, companies are deploying submerged infrastructure in the South China Sea. Australia has every geographic and technical advantage to lead in this space. What we lack is the will and vision.
Wardan Core aims to provide both.
🔧 Modular, Maintainable, and Measurable
We're not proposing utopia, we're proposing containers.
Each Wardan Core pod could be based on standardised shipping-container dimensions, retrofit or purpose-built, housing racks of AI optimised compute. Passive cold-water circulation systems would replace the massive air and freshwater cooling infrastructure currently needed for land-based centres.
Initial concepts suggest:
Deployments in 20- or 40-foot modules
Fibre connectivity routed through shore-based network exchange points
Maintenance through retrieval and swap-out, not in-situ servicing
Power sourced via offshore wind hubs or submarine cable integration with grid-connected renewables
🧠 Why Now?
AI growth is outpacing infrastructure readiness
Australia lacks sovereign AI hardware infrastructure
Offshore wind is finally getting government attention (particularly in WA, SA, and VIC)
Data centre water usage is becoming a global concern, especially in drought-prone nations like ours
Deploying offshore, modular infrastructure is not just smarter, it’s greener, cooler, and cheaper over time.
🛡️ Regulatory Realism
We are under no illusion that you can simply tow a data centre out to sea and plug it in.
Environmental approvals, seabed leasing, fisheries impact assessments, cable landing rights, naval and defence awareness, and marine zoning laws will all apply. It will require multiple departments at both state and federal levels to coordinate. But we believe this process should start now, before the pressure to “catch up” leads to less sustainable choices.
We’re initiating outreach across agencies, including:
WA’s Department of Jobs, Tourism, Science and Innovation (who have acknowledged the idea as worthwhile)
CSIRO and the Pawsey Supercomputing Research Centre (who were supportive but unable to act)
Other state and federal departments will be contacted in the coming weeks.
So far, WA has responded with polite interest but no commitment. We hope other states are hungrier.
🌏 Not Just WA
While Western Australia has an unmatched coastline for deepwater cooling, especially off the Southern Ocean, the Wardan Core model is equally applicable in:
Victoria’s Bass Strait
Tasmania’s deep cold shelf
South Australia’s southern coastline, currently exploring new tech-led clean industry models
Even parts of the NSW coast, if paired with smart zoning
This isn’t a local play, it’s a national infrastructure opportunity. One that intersects clean energy, AI capability, export potential, and sovereign data control.
💰 What Comes Next?
This is not yet a company. There is no commercial entity pitching for funding. This is a concept being brought into the public domain deliberately, so that it can be claimed, tested, interrogated, and refined by those with the means to build it.
However, paths forward could include:
Academic pilot projects
State or federal green-tech grants
CSIRO’s ON Prime or related accelerator arms
Partnerships with supercomputing centres
Private VC aligned with ESG infrastructure
What we’re seeking now is attention. Eyes. Questions. Dissent. Validation. Collaboration.
🧭 Conclusion
Australia will build new AI infrastructure. That’s a given.
The question is where, how, and who benefits.
Do we repeat the mistakes of the fossil fuel economy, onshore, centralised, grid-straining, and unsustainable, or do we seize the chance to do something bold, beautiful, and technically sound?
Wardan Core is an open invitation to rethink what’s possible.
We’re not building dreams. We’re building containers for them.
If you’re part of this conversation, or think you should be, get in touch.
📧 edward.e.shepherd@gmail.com