Australia's Megaport Secures Four New AI Infrastructure Contracts

Megaport, an Australian network-as-a-service provider, has made a substantial leap into the artificial intelligence infrastructure market by securing four new contracts with U.S.-based technology companies. These agreements, collectively valued at around A$458.9 million (approximately $329.49 million USD), are set to commence in the first half of 2027. To support this ambitious expansion, Megaport is launching a fully underwritten entitlement offer aiming to raise A$827.3 million (approximately $594 million USD). The core of these new contracts involves significant capital expenditure, estimated at nearly A$369.5 million, primarily allocated for acquiring high-performance Nvidia GPUs, alongside necessary network and storage infrastructure. Megaport's strategy includes establishing a globally distributed AI inference cloud, which will be anchored by an on-demand GPU pool backed by a A$350 million investment. This new offering will cater to enterprise customers through both contracted and consumption-based pricing models. Megaport CEO Michael Reid emphasized the immense opportunity in AI inference, stating that seamless access to GPUs, CPUs, storage, and connectivity is crucial as AI adoption accelerates. The company is betting on a surge in demand for GPU-based compute as enterprise AI shifts from model training to more latency-sensitive inference workloads. With its existing network of over 1,100 data centers across 31 countries, Megaport believes it is well-positioned to deliver AI compute closer to end-users, thereby addressing critical bottlenecks such as power, connectivity, and access to high-performance GPUs. The company utilizes both Nvidia and AMD chips in its infrastructure.