Nvidia Plans 2-Gigawatt AI Infrastructure Buildout in Australia

Nvidia expands AI factory strategy to Australia

Nvidia disclosed on September 9 that it will work with Australian cloud partners and infrastructure providers to deploy up to 2 gigawatts of AI computing capacity by 2027. The initiative will secure land, power, and data center facilities designed to house multiple generations of the company's DSX AI factory infrastructure.

According to details first reported by AI Watch, the buildout represents more than a straightforward hardware sale. Nvidia will supply its DSX platform alongside accelerated computing systems, networking equipment, software tools, and ecosystem support to the emerging network of facilities. The DSX platform integrates with Nvidia's CUDA development environment, which has become the dominant software layer for AI training and inference workloads.

The Australian expansion positions Nvidia to capture recurring revenue as enterprises, research institutions, government agencies, and startups build applications on its architecture. Rather than selling GPUs into independently operated data centers, the company is increasingly shaping the infrastructure stack from the ground up.

Why it matters

Nvidia's shift from component supplier to infrastructure platform provider changes its competitive position and revenue model. By embedding its software ecosystem and reference architectures into purpose-built AI facilities, the company can deepen customer lock-in and create demand for complementary products beyond GPUs—including CPUs, networking gear, and software licenses. For organizations evaluating AI infrastructure investments, this vertical integration means vendor decisions have longer-term architectural implications than traditional hardware purchases.

Part of a global infrastructure push

Australia joins a growing list of regions where Nvidia is deploying AI factory capacity. In Japan, the company is developing a 140-megawatt facility that will use its upcoming Rubin GPUs and Vera CPUs for robotics, manufacturing, healthcare, and digital twin applications. Across Europe, Nvidia is collaborating with partners on new AI supercomputers in multiple countries.

Each facility creates demand not only for graphics processors but for the company's broader product portfolio. As governments and enterprises seek domestic or regional AI computing resources—driven by data sovereignty requirements, latency concerns, and supply chain considerations—Nvidia's integrated approach positions it to capture a larger share of infrastructure spending.

The strategy reflects a broader evolution in Nvidia's business model. The company now offers a full-stack platform that spans silicon, networking, operating software, and application frameworks. This vertical integration gives Nvidia influence over how AI infrastructure is designed and operated, potentially extending its competitive advantage beyond the chip layer where rivals like AMD and Intel compete most directly.

Details of the Australian buildout were first reported by AI Watch.