NVIDIA is partnering with Australian cloud and data-center providers to build up to 2 gigawatts of AI computing capacity by 2027, expanding local access to the infrastructure needed to develop and run advanced AI. (Source: Image by RR)

Australian AI Infrastructure Could Support Demand Across Asia-Pacific

NVIDIA is partnering with a group of Australian cloud and infrastructure providers to build up to 2 gigawatts of AI computing capacity by 2027, expanding the country’s ability to train and operate increasingly demanding artificial intelligence systems. Partners including Firmus, Sharon AI, IREN, Megaport, ResetData, CDC, NEXTDC and AirTrunk will provide and operate the data-center infrastructure, while NVIDIA supplies its DSX platform, accelerated computing, networking, software and ecosystem support. The resulting facilities are designed to support multiple generations of NVIDIA hardware rather than a single wave of AI deployment.

The expansion, as noted at nvidianews.nvidia.com, reflects the growing physical demands behind advanced AI. NVIDIA’s DSX architecture combines computing, networking, software and facility designs into what the company calls an “AI factory,” with operators supplying the land, electricity, cooling and buildings required to house increasingly dense computing systems. The company argues that building more capacity inside Australia will give local startups, universities, researchers and enterprises access to high-performance computing while also supporting regional and global demand for AI infrastructure.

Several projects are already taking shape across the country. Sharon AI plans to deploy as many as 68,000 NVIDIA GPUs, while IREN is applying the DSX architecture to its 800-megawatt Bundey campus in South Australia. CDC says it operates more than 550 megawatts of capacity across Australia and New Zealand with another 800 megawatts under construction, while NEXTDC and AirTrunk are developing high-density facilities using technologies including liquid cooling. Together, the projects demonstrate how the AI infrastructure race is expanding beyond chips into electricity, networking, cooling, real estate and large-scale data-center construction.

NVIDIA also wants the infrastructure to support an Australian AI ecosystem rather than simply provide overseas computing capacity. Local organizations will gain greater access to NVIDIA’s Nemotron open models and accelerated computing, with companies including healthcare AI developer Heidi and enterprise software maker Atlassian already using NVIDIA technology to develop specialized applications. The strategy connects locally available compute with models and software, positioning Australia’s infrastructure expansion as both a regional source of AI capacity and a foundation for companies building AI products inside the country.

read more at nvidianews.nvidia.com