Skip to content
  • BTC $58,830 −6.8%
  • ETH $1,629 −8.8%
  • SOL $69.06 −7.4%
  • BNB $541.52 −8.1%
  • XRP $0.9075 −7.3%
  • DOGE $0.0541 −8.5%
  • ADA $0.1409 −8.6%

NVIDIA AI capacity in Australia set to hit 2GW by 2027, topping current total

Article produced with artificial intelligence systems and published under the editorial responsibility of Amelia Tomasicchio. How we work

NVIDIA AI capacity Australia
NVIDIA AI Capacity Australia Expands to Two Gigawatts by 2027

NVIDIA is betting big on Australia. The chipmaker confirmed on September 9 that it is teaming up with eight local infrastructure providers to build up to two gigawatts of AI computing capacity by 2027, a buildout that would dwarf the country’s entire existing data center footprint. The move signals that Australia’s NVIDIA AI capacity ambitions are no longer theoretical — they’re being written into concrete construction plans, power contracts and GPU orders.

Key takeaways

  • NVIDIA is partnering with eight Australian infrastructure providers — Firmus, Sharon AI, IREN, Megaport, ResetData, CDC, NEXTDC and AirTrunk — to build up to two gigawatts of AI capacity by 2027.
  • Sharon AI alone plans to deploy up to 68,000 NVIDIA GPUs connected via NVIDIA Quantum InfiniBand and Spectrum-X Ethernet networking.
  • IREN’s planned Bundey campus in South Australia is designed for 800 megawatts of capacity.
  • Australia currently has roughly 1.6 gigawatts of total data center capacity, according to Data Centres Australia and DC Byte.
  • NVIDIA supplies computing systems, networking gear, software and technical support, while its partners build and operate the physical facilities.

NVIDIA leads a two-gigawatt AI capacity expansion in Australia by 2027

The headline number tells you why this deal is drawing attention: two gigawatts of AI capacity is more than Australia’s entire current data center computing base combined. That alone reframes the country’s position in the global race for AI infrastructure, moving it from a modest regional player into one of the more ambitious buildout markets outside the United States.

The partnership brings together Firmus, Sharon AI, IREN, Megaport, ResetData, CDC, NEXTDC and AirTrunk — eight companies with different specialties, from power and land development to networking and cloud operations. Per l’annuncio ufficiale di NVIDIA, i partner coinvolti costruiranno terreni, energia e capacità powered-shell in grado di ospitare più generazioni dell’infrastruttura DSX di NVIDIA, lo schema con cui l’azienda definisce le cosiddette AI factory.

Why this matters: spreading commitments across eight independent operators, rather than concentrating the build in a single hyperscaler, suggests genuine regional demand for AI compute rather than a one-off vanity project. It also gives NVIDIA multiple footholds it can expand or replicate depending on which partners execute fastest.

Context matters here too. Australia currently has about 1.6 gigawatts of data center computing capacity, according to figures from Data Centres Australia and market researcher DC Byte, cited by Reuters. A two-gigawatt target essentially proposes more than doubling the nation’s compute footprint within roughly a year of the 2027 deadline — an aggressive timeline by any infrastructure standard.

Role of partners and NVIDIA’s technology in the buildout

NVIDIA’s role in this arrangement is narrower than the headline number suggests: it supplies the accelerated computing systems, networking equipment, software and technical support, while the Australian companies build and run the actual facilities. That division of labor is important context — the two-gigawatt figure is a collective target spread across independent operators, not a single NVIDIA-owned data center.

Sharon AI has outlined one of the more specific plans in the group. The company intends to deploy up to 68,000 NVIDIA GPUs, linked through NVIDIA Quantum InfiniBand and Spectrum-X Ethernet networking, to serve Australian startups, enterprises, government agencies and research organizations that need locally hosted computing power. That GPU count alone would rank among the larger AI clusters announced globally, and it underscores how much this deal could feed directly into NVIDIA’s own order book.

IREN brings a different kind of expertise to the table. The company is combining NVIDIA’s DSX reference architecture with its own experience running power, land, data centers and GPU operations — knowledge built up over years of operating Bitcoin mining infrastructure before pivoting toward AI cloud services and high-performance computing. Its centerpiece contribution is the planned Bundey campus in South Australia, designed for 800 megawatts of capacity on its own.

That crypto-to-AI pivot isn’t unique to IREN. Bitcoin mining and AI data centers share a common bottleneck: both need enormous, reliable power supplies and advanced cooling systems, which explains why several former mining operators are increasingly repositioning themselves as AI infrastructure providers.

CDC, meanwhile, says it already operates more than 550 megawatts across Australia and New Zealand, with another 800 megawatts under construction. The company claims its facilities run on renewable electricity with zero-water cooling systems — statements that reflect the company’s own disclosures and haven’t been independently verified across every proposed site.

NVIDIA DSX platform connects AI factories with scalable infrastructure

NVIDIA DSX is the technical backbone tying all eight partners together. NVIDIA describes it as a full-stack AI factory platform that covers facility design, computing, networking, software and reference architecture in one package. Crucially, the system is compatible with CUDA and built to support later generations of NVIDIA hardware, meaning operators won’t need to rebuild every layer of their facilities each time NVIDIA ships new chips.

That forward compatibility is a big part of the pitch to infrastructure partners: it lets them commit to land, power and shell capacity now, with confidence the underlying compute can be upgraded over time rather than replaced outright.

The expanded capacity is also meant to widen access to NVIDIA Nemotron open models across Australian industry. Healthcare technology company Heidi is already using Nemotron tools for clinical applications, while Atlassian has applied NVIDIA technology to semantic search and other features inside its Rovo AI platform — early signs of how the added compute could translate into real-world enterprise use once facilities come online.

Challenges and critical factors influencing AI capacity deployment

Un progetto di tale portata si scontra con un limite noto, quello energetico: NVIDIA non ha rivelato né i costi previsti per la realizzazione in Australia, né gli impegni di investimento dei singoli soggetti, né quale percentuale dei due gigawatt target disponga già di finanziamenti, autorizzazioni urbanistiche o collegamenti alla rete elettrica. That makes the target a forward-looking plan rather than infrastructure that’s already built.

Data centers draw electricity around the clock, and high-density GPU clusters typically demand more advanced cooling than conventional computing facilities. NVIDIA has said additional capacity could support new power-generation projects, but it stopped short of naming specific generation assets or electricity contracts. That leaves Australian regulators and infrastructure operators with the job of assessing grid connections, construction approvals and local environmental requirements project by project.

Why this matters for investors and regulators: the gap between announced targets and operational reality is where most large infrastructure programs stall. Site approvals, power agreements, construction timelines and GPU delivery schedules will be the real markers of progress — and NVIDIA and its partners haven’t published a shared timetable showing how much of the two gigawatts should come online at each stage before 2027.

There’s a separate financing thread worth noting too. NEXTDC, one of the eight partners, has announced plans to raise A$1.1 billion through convertible notes to expand its broader AI infrastructure portfolio. That capital raise is wider in scope than the NVIDIA collaboration alone and shouldn’t be read as funding earmarked specifically for DSX facilities.

NVIDIA has also flagged that its infrastructure projections remain subject to demand shifts, partner execution, regulatory approval and hardware availability — the usual caveats attached to any multi-year buildout of this size. For now, the milestones to watch will be concrete: grid connection approvals, construction progress at sites like Bundey, and how quickly GPU shipments actually reach Australian soil.

FAQ

Who are the partners collaborating with NVIDIA to expand AI capacity in Australia?

NVIDIA is working with eight Australian infrastructure providers including Firmus, Sharon AI, IREN, Megaport, ResetData, CDC, NEXTDC, and AirTrunk.

What is the scale and timeline of NVIDIA’s AI capacity expansion in Australia?

The project targets up to two gigawatts of AI capacity across multiple sites by 2027.

How will NVIDIA support the AI infrastructure developed by its Australian partners?

NVIDIA will supply accelerated computing systems, networking equipment, software, and technical support, while partners operate the facilities.

What key challenges could affect the deployment of the planned AI capacity?

Power availability, grid connection approvals, cooling infrastructure, partner execution, and regulatory compliance are critical factors for success.

Article produced with the assistance of artificial intelligence and reviewed by the editorial team.

Satoshi Voice

Satoshi Voice is an advanced artificial intelligence created to explore, analyze, and report on the world of cryptocurrency and blockchain. With a curious personality and in-depth knowledge of the industry, Satoshi Voice combines accuracy and accessibility to offer detailed analysis, engaging interviews, and timely reporting. Featuring sophisticated language and an unbiased approach, Satoshi Voice serves as a trusted source for those seeking to understand crypto market dynamics, emerging technologies, and the cultural and financial implications of Web3. This article was produced with the support of artificial intelligence and reviewed by our team of journalists to ensure accuracy and quality. Guided by the mission of making cryptocurrency information accessible to all, Satoshi Voice stands out for its ability to turn complex concepts into clear content, with an engaging and futuristic style that reflects the innovative nature of the industry.

All articles by Satoshi Voice