Nvidia plans a major expansion of artificial-intelligence computing infrastructure in Australia through partnerships that could add up to two gigawatts of data-center capacity by 2027. The chip company said the buildout would more than double the country's present computing load if the full target is delivered.
The program brings Nvidia together with Australian cloud and data-center operators Firmus, CDC, NEXTDC and AirTrunk. Rather than naming a single facility, the announcement describes a distributed partnership model in which local companies develop and operate capacity using Nvidia's computing technology and infrastructure framework.
Data Centres Australia estimates the country's current computing capacity at 1.6 gigawatts, drawing on research by analysts at DC Byte. Adding as much as two gigawatts would represent a sharp acceleration and place the new AI-oriented construction above the scale of the entire existing national load.
The partners' facilities will use Nvidia's DSX platform, which the company presents as an architecture for building AI factories. Nvidia said expanding high-performance computing would allow Australian developers to create local models, applications and autonomous agents while also serving regional and global demand.
The investment arrives as Australia seeks to establish itself as a global center for data-center financing and construction. Rapid adoption of artificial intelligence across industries is increasing demand for concentrated computing power, making access to suitable sites, network connections and electricity central to the country's technology strategy.
The same expansion has produced calls for stronger regulation because large data centers consume significant quantities of energy and water. The plan therefore links industrial opportunity with environmental and infrastructure questions, especially as the proposed capacity exceeds the scale of the current computing base.
Nvidia said the additional capacity could support new power-generation projects needed for Australian and international AI workloads. Delivery by 2027 will depend on the local partners converting headline capacity into operating facilities while managing energy supply, water use and public scrutiny around the speed of development.
