SharonAI lines up a six-year NVIDIA-backed plan to expand Australia’s AI data-center capacity

SharonAI Holdings has entered into a six-year artificial intelligence infrastructure collaboration with NVIDIA, setting out a plan to expand data-center and GPU capacity in Australia, the company said in a report carried by Yahoo Finance.

The agreement is described as focusing on increasing compute availability for AI workloads, a market need that has intensified as companies seek local capacity to run training and inference. For NVIDIA, the arrangement fits its broader push to expand the ecosystem of data-center platforms built around its GPU and AI software stack, while giving infrastructure partners a path to deploy capacity tied to demand.

SharonAI, which trades in the U.S. as SHAZ, is positioning the collaboration as a way to reshape its operating footprint. The reported terms also include revenue sharing, though the article summary did not detail which revenue streams are shared, how costs are allocated, or what performance milestones, if any, govern payments.

Under the reported structure, SharonAI is expected to use the NVIDIA collaboration to support additional data-center and GPU capacity in Australia over the life of the deal. These capacity expansions typically involve not only processors and servers, but also power, cooling, networking, and hosting operations, all of which can drive the economics of AI cloud offerings.

The report frames the deal as an effort to improve the economics of providing AI infrastructure, implying that pairing long-term hardware and technology support with revenue-linked commercial terms could reduce uncertainty compared with spot leasing or short-horizon procurement. For customers, the potential upside is more predictable access to GPU resources, though the timing and scale will depend on how quickly new capacity comes online.

NVIDIA has previously emphasized that AI infrastructure is increasingly built as end-to-end systems that combine hardware, accelerated computing software, and deployment tooling. While the Yahoo summary does not cite specific NVIDIA products named in the deal, the collaboration concept generally points to NVIDIA’s data-center platform approach, in which its GPUs and software are meant to work together for training and inference.

Still, the details that investors and customers would likely scrutinize are not included in the report summary. It does not specify the expected number of GPUs, the location(s) and timeline for new data-center additions in Australia, the precise definition of “revenue sharing,” or any minimum commitments by either side.

What to watch next is whether SharonAI provides further disclosures, such as capital expenditure plans, capacity targets, contract durations with end customers, and any milestones tied to the NVIDIA collaboration. Additional clarity on pricing, cost structure, and when the expanded capacity will be available could determine how quickly the deal translates into revenue.