Nvidia has partnered with eight Australian cloud and data centre providers to develop up to two gigawatts of AI infrastructure by 2027, as demand for high-performance computing surges.

Firmus, Sharon AI, IREN, Megaport, ResetData, CDC, NEXTDC and AirTrunk will provide the land, power, facilities and operations required to host multiple generations of Nvidia DSX infrastructure.

Nvidia will supply its full-stack DSX platform, including accelerated computing hardware, networking, software and reference designs.

The resulting “AI factories” are intended to give Australian startups, universities, researchers, government agencies and enterprises greater access to high-performance computing without relying entirely on offshore infrastructure.

“Through our partnerships with Australian NCPs, this infrastructure can support multiple generations of Nvidia compute and expand local access to accelerated computing,” Nvidia vice-president Raj Mirpuri said.

Sharon AI plans to deploy up to 68,000 Nvidia GPUs connected through Quantum InfiniBand and Spectrum-X Ethernet networking.

AI Factories Are Redefining Data Centers, Enabling Next Era of AI | NVIDIA Blog

IREN will use Nvidia’s architecture at its proposed 800MW Bundey campus in South Australia, combining data centres, power infrastructure and GPU operations.

Firmus is expanding its Project Southgate AI infrastructure initiative, while Megaport will provide accelerated computing through subsidiary Latitude.sh and connect it to the company’s global software-defined network.

ResetData is targeting organisations requiring Australian data residency, including enterprises, governments and researchers.

CDC currently operates more than 550MW of capacity across Australia and New Zealand, with another 800MW under construction. Its facilities use renewable electricity and zero-water cooling technology designed for high-density computing.

NEXTDC and AirTrunk are also expanding liquid-cooled facilities capable of supporting increasingly power-intensive AI systems.

The companies have not detailed the total investment required or precisely how much of the proposed two-gigawatt capacity will be operational by the end of 2027.

Nvidia says the expansion will also provide wider access to its Nemotron open AI models. Australian companies Heidi and Atlassian are already using the technology for clinical AI and enterprise search applications respectively.