Iris Coleman
Sep 10, 2026 00:37
NVIDIA partners with key Australian firms to build AI infrastructure with a 2-gigawatt capacity by 2027, boosting AI innovation and sovereignty.
NVIDIA (NASDAQ: NVDA) is teaming up with Australian data center operators and cloud providers to construct a 2-gigawatt AI infrastructure network by 2027. This initiative, leveraging NVIDIA’s DSX AI factory platform, aims to meet escalating demand from startups, researchers, and enterprises building advanced AI applications locally.
The collaboration includes major players like Sharon AI, Firmus, ResetData, CDC, NEXTDC, and AirTrunk. These partners will provide land, power, and cooling infrastructure, while NVIDIA supplies its full-stack DSX platform, GPUs, networking solutions, and software support. The project is intended to enable Australian companies and institutions to develop and host AI models without outsourcing to global providers, ensuring data sovereignty and reducing latency for critical applications.
“AI factories turn energy into intelligence—an essential resource for the AI economy,” said Raj Mirpuri, NVIDIA’s VP of Global AI Clouds and Infrastructure. The DSX platform, compatible with NVIDIA’s CUDA ecosystem, supports multiple generations of compute infrastructure, offering durability and efficiency for long-term deployments.
Strategic AI Expansion Amid Global Competition
Australia’s established renewable energy capacity and digital infrastructure make it a prime location for NVIDIA’s expansion. The country’s reputation for stable governance and reliable energy further strengthens its appeal for AI-native startups and enterprises.
This project aligns with NVIDIA’s broader AI infrastructure strategy, where it has recently announced similar initiatives globally. On August 26, 2026, NVIDIA partnered with AWS to deploy 2 million GPUs across AWS’s global infrastructure. Earlier in August, NVIDIA guaranteed exclusive AI infrastructure for SB Energy’s PORTS-Pike site in Ohio, with OpenAI expected as a customer. These moves highlight NVIDIA’s intent to dominate not only in chip manufacturing but also in end-to-end AI infrastructure solutions.
Key Partners and Local Impact
Sharon AI, one of the lead partners, plans to deploy 68,000 NVIDIA GPUs across its facilities, using NVIDIA’s Quantum InfiniBand and Spectrum-X Ethernet networking for high-performance compute needs. Meanwhile, Firmus is expanding its Project Southgate initiative to support hyperscale demand across Australia and the broader Asia-Pacific region.
Other collaborators like ResetData and Megaport are focused on providing AI-ready infrastructure with strong data residency assurances, critical for government and healthcare clients. These partnerships aim to position Australia as a regional hub for AI innovation, supporting industries from healthcare to enterprise software.
Market Context
NVIDIA’s stock (NVDA) traded at $223.67 as of September 10, 2026, with a slight decline of 0.87% over the past 24 hours. The company’s market cap stands at $5.43 trillion, reflecting its firm grip on the AI and semiconductor sectors. Investors should note that NVIDIA’s growing focus on sovereign AI infrastructure could offset near-term chip demand fluctuations by creating long-term recurring revenue streams from its DSX platform and cloud partnerships.
What’s Ahead
The Australian buildout is scheduled to reach full capacity by 2027, with immediate benefits for local organizations like Atlassian and Heidi, which are leveraging NVIDIA’s Nemotron models to develop industry-specific AI solutions. With NVIDIA’s continued investments in infrastructure and its reputation for delivering cutting-edge compute capabilities, Australia could emerge as a significant player in the global AI economy.
For traders, NVIDIA’s ongoing expansion into sovereign and regional AI infrastructure offers a glimpse into its long-term growth strategy, which goes beyond hardware sales to encompass software, services, and strategic partnerships. This diversified approach could provide resilience against cyclical downturns in the semiconductor market.
Image source: Shutterstock
