NVIDIA Partners With Eight Australian Firms to Build Up to 2-Gigawatt AI Infrastructure Network by 2027

NVIDIA has announced a sweeping expansion of artificial intelligence computing infrastructure across Australia, partnering with eight data center and cloud providers to build out capacity designed to host multiple generations of AI factory hardware, with a target of up to 2 gigawatts of total power by 2027.

The company is collaborating with Firmus, Sharon AI, IREN, Megaport, ResetData, CDC, NEXTDC and AirTrunk, all of which are either NVIDIA Cloud Partners or AI infrastructure partners. The arrangement positions NVIDIA as the technology supplier while the participating companies operate the physical infrastructure, with NVIDIA delivering its DSX platform, accelerated computing, networking, software and ecosystem support.

The DSX Platform at the Center of the Buildout

The entire initiative is built around NVIDIA's DSX platform, which the company describes as a full-stack AI factory system spanning facilities infrastructure, computing, networking, software and reference designs.

NVIDIA says the platform is compatible with its CUDA ecosystem and is designed to be enhanced through software updates over the lifespan of the physical infrastructure. According to NVIDIA, the DSX architecture is intended to help operators serve a broad range of workloads and customers, improve performance and utilization, and extend the useful life of their assets.

The company is pitching the infrastructure model as a new investable asset class, using terms including productivity, fungibility, durability and rentability to characterize the long-term economics of the platform. Raj Mirpuri, vice president of global AI clouds and infrastructure ecosystem at NVIDIA, said the company views AI factories as systems that turn energy into intelligence.

He described the DSX platform as capable of supporting multiple generations of NVIDIA compute, which he said would expand local access to accelerated computing and NVIDIA's Nemotron open models for Australian developers and organizations.

Partner by Partner: What Each Company Is Building

Each of the eight partners is pursuing a distinct role within the broader buildout. Firmus is expanding its Project Southgate initiative, which involves developing NVIDIA-powered AI factories designed to support hyperscale and AI-native demand across Australia and the Asia-Pacific region.

Oliver Curtis, cofounder and co-CEO of Firmus, said the project is intended to give Australian businesses, researchers and innovators access to computing capacity locally while positioning Australia to serve growing AI demand across the broader region. Sharon AI is deploying up to 68,000 NVIDIA GPUs connected with NVIDIA Quantum InfiniBand and NVIDIA Spectrum-X Ethernet networking, using the DSX platform to serve startups, enterprises and research organizations.

James Manning, cofounder and CEO of Sharon AI, said the company is aiming to bring gigawatt-scale AI factories to Australian enterprises, researchers, government agencies and AI-native companies, citing sovereignty and security as key considerations in its design choices. IREN is combining NVIDIA's DSX AI factory architecture with what it describes as its own vertically integrated platform spanning power, land, data centers, GPU deployment and infrastructure operations.

Daniel Roberts, cofounder and co-CEO of IREN, said the partnership creates a repeatable blueprint to bring capacity online at scale, and that IREN is applying that blueprint across its development portfolio. Megaport, ResetData, CDC, NEXTDC and AirTrunk are also named as participating partners in the collaboration, though the announcement does not provide detailed breakdowns of each company's specific deployment plans or capacities.

Who the Infrastructure Is Designed to Serve

NVIDIA identified a broad range of potential users for the expanded infrastructure, including Australian startups, universities, researchers, enterprises, developers and AI-native companies. The announcement also named two existing organizations as examples of entities that would benefit from expanded access to NVIDIA accelerated computing and Nemotron open models: Heidi and Atlassian. Both are described as organizations that will be able to build regional AI models, applications and agents for industries nationwide.

The announcement frames Australia's existing strengths as foundational to the initiative's viability. NVIDIA cited the country's established sustainable energy and digital infrastructure, deep technology talent, and its reputation as a trusted global economic partner as reasons why Australia provides a strong foundation for training, deploying, and scaling AI.

Scale and Strategic Positioning

The 2-gigawatt figure cited in the announcement represents the upper range of the planned capacity buildout across all participating partners by 2027. NVIDIA describes the expansion as intended not only to serve domestic Australian demand but also to support regional and global demand for AI compute.

The initiative reflects a broader pattern in which NVIDIA is working to establish AI factory infrastructure in key markets by partnering with local data center operators and cloud providers rather than building and operating the facilities itself.

Under this model, the local partners provide the land, power, and physical shell while NVIDIA supplies the technology stack that runs inside them. The September 9, 2026 announcement did not specify individual power commitments from each of the eight partners, nor did it detail the financing arrangements or timelines specific to each company's deployment beyond IREN's description of applying the DSX blueprint across its development portfolio and Sharon AI's disclosure of its GPU count target.