Policy
Examine the evidence behind the development pipeline, its stages and the gaps that matter to infrastructure decisions.
ABOUT THE ATLAS
A clearer picture of the infrastructure behind AI.
ANZ Compute Atlas is a public research initiative of Accelerate AI Partners. We focus on Australia’s data centre and compute landscape, with New Zealand covered alongside it.
Why we started the atlas.
The data centre and compute story is scattered across operator websites, planning documents, investor announcements and industry reports. Details can be difficult to compare, incomplete or conflicting. Even a simple question about how much capacity is operating can produce different answers, depending on what is being counted.
We started ANZ Compute Atlas to bring that evidence into one accessible reference. It connects facility records, disclosed capacity and development milestones with their original sources, so readers can see what a figure means and where it came from.
The Atlas brings an authoritative, open and citable view to Australia and New Zealand’s compute infrastructure. Explore the landscape through the map, table, timeline and research, with clear definitions and original sources behind the figures.
For people shaping the conversation.
Media
Follow a capacity figure to the facility, physical scope, date and original source behind it.
Investors
Understand capacity, demand and delivery risk, and examine the evidence behind announced projects and commercial commitments.
The AI community
Understand the difference between electrical capacity, a data centre’s development status and evidence of usable AI compute.
THE BUSINESS BEHIND THE ATLAS
An initiative of Accelerate AI Partners.
Accelerate AI Partners helps mid-sized organisations adopt generative AI through hands-on, in-person training and practical advisory. Its work focuses on applying AI to business needs.
The atlas is part of that business’s public research and analysis. Its purpose is to make the infrastructure conversation more informed, with evidence and limitations available for readers to inspect.
THE PEOPLE BEHIND THE BUSINESS
Meet the Accelerate AI Partners team.
Our partners bring experience across engineering, strategy, transformation, finance and governance, with a shared focus on helping organisations put AI to work.
Andrew Gargiulo
Founder & Director
Andrew brings 15 years of experience across engineering, strategy consulting and corporate transformation, spanning five continents. He founded Accelerate AI Partners to help mid-sized organisations turn AI into faster delivery, better quality and growth through practical training and straight-talking advisory.
Samantha Taylor
Partner
Samantha is an engineer, futurist and strategic leader with 30 years across consulting, research and development, government and private enterprise, including Transurban. She connects technical expertise with business vision, helping leaders and teams turn AI and data into practical, enduring outcomes.
John Johansson
Partner
John is a Chartered Accountant with over 15 years of international experience across consulting, finance, compliance and operational leadership, including a decade at the ASX. He helps organisations simplify operations, strengthen governance and adopt AI with confidence.
Meet the team and explore our work at Accelerate AI Partners .
Transparency is part of the product.
Sources, definitions and research gaps are published alongside the directory. Capacity claims retain their electrical basis, development stage and physical scope. Unknown figures remain unknown, and a future target is not treated as proof that a facility has opened.
A megawatt measures electrical power. It does not establish GPU quantity, accelerator performance, utilisation or capacity available to purchase. More detailed compute research needs evidence of the equipment, its location, deployment status and access.
Our source register connects government authorities, operator disclosures and industry research to the claims they support. The methodology explains how records are counted and how to interpret the figures.
Use and cite the Atlas.
Figures are dated to dataset 2026-11-03-launch-13. Capacity subtotals cover documented, comparable claims; coverage varies across sites and operators. They should not be read as a complete national capacity total or installed GPU inventory.
The coverage page shows which locations and capacity claims have supporting evidence. Check the source dates when using a figure in a report or decision.
To cite the atlas, include the relevant facility profile, release identifier, claim and underlying source. If you can improve a record, provide the source and the specific scope or date it establishes.