Sharon AI has secured a five-year, US$373 million cloud computing agreement with a global artificial intelligence platform, giving the Australian neocloud a major contracted customer commitment as demand for AI infrastructure accelerates. Sharon AI Holdings Inc. and its subsidiaries announced the agreement on August 4, positioning the deal as a significant expansion of its commercial footprint and a further step in building large-scale AI infrastructure in Australia. The contract carries a total value of US$373 million across its five-year term, with the company expecting revenue from the agreement to begin in the first quarter of 2027. For Sharon AI, the more important signal sits beyond the headline contract value: the agreement converts a substantial portion of its planned AI infrastructure into contracted capacity tied to an identified customer.

The company expects to deploy cloud computing solutions under the agreement across its AI infrastructure in Australia, keeping the investment and compute deployment within a market that is increasingly seeking greater domestic access to advanced AI capacity. Sharon AI’s total AI Factory capacity remains at 132 megawatts following the agreement, while contracted capacity to end customers has climbed to 120MW. That leaves the company with a much smaller portion of its planned power footprint available for future customer commitments, increasing the strategic importance of how it allocates remaining capacity. The company also plans to increase the GPU count across its AI Factory platform from 62,000 to 64,000 NVIDIA GPUs by the middle of 2027.

2,048 Blackwell Ultra GPUs Start The Deployment

The first deployment under the new agreement is expected to use 2,048 NVIDIA Blackwell Ultra B300 GPUs, giving the contract an immediate connection to one of the most advanced compute platforms in Sharon AI’s planned infrastructure portfolio. The initial deployment represents only the starting point of the five-year relationship, rather than a ceiling on the amount of infrastructure the customer could consume. Additional deployments may take place throughout the agreement based on the customer’s requirements and the terms established under the contract. That structure gives Sharon AI room to expand the infrastructure supporting the customer as its AI workloads evolve.

The distinction matters in an AI infrastructure market where customers increasingly need access to large pools of accelerator capacity rather than isolated GPU deployments. A long-duration cloud agreement can give an infrastructure provider greater visibility into future utilization while allowing the customer to secure access to compute without building and operating the underlying physical infrastructure itself. Sharon AI’s move therefore links its GPU expansion directly with contracted demand instead of treating accelerator growth purely as a speculative capacity bet. Meanwhile, the 132MW AI Factory platform provides the power envelope within which the company expects to scale its infrastructure and serve additional AI customers.

Australia’s Sovereign AI Capacity Gets A Boost

The agreement also places Sharon AI’s infrastructure strategy within Australia’s broader push to establish itself as an important destination for AI infrastructure investment across the Asia-Pacific region. Deploying high-performance compute inside Australia can give international AI platforms access to advanced infrastructure while supporting the development of a wider domestic technology ecosystem. The company expects its infrastructure expansion to create opportunities for Australian businesses, researchers and government organizations seeking access to high-performance AI compute closer to their operations. That creates a two-sided proposition for the country: Australia can attract global AI workloads while simultaneously expanding local access to the infrastructure required to develop and run them.

For Sharon AI, the contract strengthens the commercial case for building large-scale AI capacity in Australia rather than treating the country simply as another location in a global cloud footprint. Long-term commitments can also improve visibility for infrastructure investment, particularly as AI workloads continue to place pressure on power availability, accelerator supply and data center capacity. Additionally, contracted demand can help anchor the economics of new infrastructure as providers commit capital to increasingly expensive GPU and power-intensive deployments. Sharon AI is now carrying 120MW of contracted end-customer capacity against its 132MW AI Factory platform, making customer commitments a central component of its expansion strategy.

Sharon AI Bets On Multi-Year AI Infrastructure Demand

The deal arrives as the economics of AI infrastructure continue shifting toward larger and longer-term commitments between compute providers and AI platforms. Training and inference workloads can require substantial accelerator fleets, while customers also need infrastructure that can scale as model sizes, usage patterns and compute requirements change. Sharon AI’s agreement provides a five-year framework within which that expansion can occur, with the company able to make additional deployments subject to customer requirements and contractual terms. The arrangement therefore gives the neocloud a path to grow its relationship beyond the initial 2,048-GPU deployment.

The company’s planned increase to 64,000 NVIDIA GPUs by mid-2027 also shows the scale of the infrastructure strategy behind the announcement. Rather than presenting the contract as a standalone cloud win, Sharon AI is folding it into a broader AI Factory platform that combines power capacity, accelerator deployment and long-term customer commitments. The remaining 12MW between its 132MW total capacity and 120MW already contracted capacity becomes strategically important as the company looks to balance new customers with the infrastructure required to support existing agreements. In that context, the US$373 million contract is both a revenue opportunity and an indicator of how quickly Sharon AI is converting its infrastructure pipeline into committed demand.

CEO Says Sovereign Compute Is Driving Demand

James Manning, Co-founder and Chief Executive Officer of Sharon AI, said, “This agreement represents an important milestone in the continued expansion of Sharon AI’s customer base and contracted AI infrastructure capacity. As organizations increasingly seek access to sovereign, high-performance AI compute, we remain focused on delivering scalable infrastructure that supports the evolving needs of AI platforms, enterprises and governments.” Manning added, “We continue to see strong demand for AI infrastructure in Australia and across the Asia-Pacific region, and this agreement reflects our strategy of securing long-term customer commitments as we expand our AI Factory platform.”

The comments underline the central thesis behind Sharon AI’s expansion: demand for AI compute is becoming increasingly connected to where infrastructure sits, how much power it can secure and whether customers can obtain reliable access over multiple years. Australia gives the company a market in which sovereign compute requirements and international AI demand can converge, while its AI Factory model gives Sharon AI a framework for scaling accelerator capacity around contracted customers. The US$373 million agreement consequently gives the company a substantial commercial anchor as it works toward its 64,000-GPU target and continues expanding its Australian infrastructure. For the wider Asia-Pacific AI market, the deal signals that sovereign compute is moving from a strategic talking point toward long-term infrastructure commitments with real capacity and capital behind them.