NVIDIA Backs 2-Gigawatt Australian AI Infrastructure Buildout by 2027
NVIDIA is partnering with eight Australian data center providers to develop up to two gigawatts of collective AI computing capacity by 2027, exceeding the nation's current data center footprint.
Key takeaways · 3
- 01
The planned 2-gigawatt collective capacity exceeds Australia's current 1.6-gigawatt data center footprint.
- 02
Sharon AI plans to deploy up to 68,000 GPUs connected via Spectrum-X Ethernet and Quantum InfiniBand.
- 03
NVIDIA will provide hardware and software, while local partners will operate the facilities.
National Capacity Expansion
NVIDIA is collaborating with eight Australian infrastructure providers to develop up to two gigawatts of AI factory capacity by 2027. [1][2]
The partnership includes Firmus, Sharon AI, IREN, Megaport, ResetData, CDC, NEXTDC, and AirTrunk. [1][2]
This collective capacity target would exceed Australia's current total data center computing capacity of 1.6 gigawatts. [2]
The Australian partners will manage facility operations, while NVIDIA will supply its DSX platform, computing systems, networking, and software. [1][2]
Specific Partner Deployments
Sharon AI intends to install up to 68,000 NVIDIA GPUs using NVIDIA Quantum InfiniBand and Spectrum-X Ethernet networking. [2]
Another partner, IREN, has designed its planned Bundey campus in South Australia to support 800 megawatts of overall capacity. [2]
These deployments aim to provide local computing resources for Australian organizations, such as Atlassian and Heidi, to develop regional AI models. [1]
What it means
The planned two-gigawatt expansion illustrates the massive physical infrastructure footprint required for sovereign AI development, effectively aiming to more than double Australia's existing data center capacity in a short timeframe. By keeping hardware operations under the purview of local providers while standardizing on the DSX architecture, NVIDIA secures regional market capture without absorbing the real estate and power management overhead. What the sources don't address: How the Australian energy grid will adapt to supply the massive power demands of this rapid two-gigawatt expansion by 2027.
The scale of this infrastructure buildout signals a significant shift toward localized, sovereign AI capabilities. By deploying state-of-the-art GPUs in regional data centers, organizations can bypass the latency and data sovereignty concerns associated with offshore cloud providers.
Why it matters
Put this to work — one session a day, built for your industry.
Create a free account for a daily session — eight questions and one real-work challenge, on the news that affects your role.
Start freeHow this developed
11 September 2026
NVIDIA Backs 2-Gigawatt Australian AI Infrastructure Buildout by 2027
11 September 2026
Event created from source cluster.