SuperX wins its first Australian order for 128 NVIDIA B300 AI server units, via Ezisight

SuperX AI Technology Limited said it has secured its first purchase order in Australia, a batch of 128 NVIDIA B300 AI server units sourced through its subsidiary channel and routed via partner Ezisight. The company framed the order as its official entry into the Australian market, marking a step in its effort to sell full-stack AI infrastructure offerings that include hardware deployment support alongside computing equipment.

In the announcement carried by Yahoo Finance, SuperX described the transaction as its “first batch” of NVIDIA B300 AI servers, specifying the unit count of 128. SuperX identified the ordering or purchasing counterparty as Ezisight and said the order was placed for delivery of the B300 servers through its subsidiary, SuperX Microinference Pte Ltd.

The reported deal is notable in part because it connects a smaller, Nasdaq-listed AI infrastructure supplier with one of the best-known names in accelerated computing. NVIDIA’s B300 servers are positioned by the industry as purpose-built platforms for running AI workloads, including model training and inference, and are typically deployed in data centers that need high-throughput compute and networking capacity.

SuperX did not, in the report, provide further detail on order timing, delivery schedule, contract value, or payment terms. It also did not disclose whether the purchase order is part of a longer master supply agreement, or whether it includes any service components such as installation, configuration, or ongoing support. Those commercial specifics remain unreported in the announcement.

For SuperX, the Australia pitch matters because it indicates the company is moving beyond broader “go-to-market” announcements into specific hardware procurement tied to a regional channel partner. Australia has become a target geography for data center buildouts and AI compute demand as organizations look to expand local capacity for AI-driven analytics, automation, and customer-facing applications, especially where latency, data residency, or compliance factors matter.

For NVIDIA, the procurement serves as another example of demand migrating through the supply chain to system integrators and infrastructure providers. Rather than selling only as individual chips, NVIDIA’s data center platforms increasingly show up inside system-level deployments assembled by resellers, integrators, and service-oriented infrastructure companies. Even without financial disclosure, orders like this illustrate ongoing end-market interest in AI server capacity built around NVIDIA architectures.

Still, investors and customers will likely want clarity on what happens next. SuperX did not state whether additional batches are expected under the same customer relationship, nor did it specify performance targets, configuration details, or expected timelines for deployment in Australia. Without those items, it is difficult to gauge how quickly the order could translate into recognized revenue or how it will affect capacity planning for future shipments.

Going forward, the key indicates to watch are whether SuperX provides follow-on updates such as delivery completion, revenue recognition, and any expansion orders from Ezisight or other Australian channel partners. For the broader market, the reaction will hinge on whether this first Australia entry becomes repeatable, suggesting durable demand for NVIDIA-based AI server capacity delivered through SuperX’s infrastructure model.