Megaport unveils Globally-Distributed AI Inference Cloud, backed by new contracts and an A$827.3M raise

About Megaport Ltd

Megaport Limited (ASX: MP1) is a global automated infrastructure platform that brings network and compute together on a single software-driven platform, enabling businesses to deploy secure, scalable infrastructure closer to users, data and clouds. The company partners with service providers, data centres and system integrators to provide programmable connectivity across more than 1,100 enabled locations in 31 countries, and is ISO/IEC 27001 certified. Its compute capabilities have been significantly expanded since the acquisition of Latitude.sh.

The AI inference opportunity

Megaport’s strategy responds to a shift in artificial intelligence workloads from training to latency-sensitive inference, where demand for GPU-based compute is currently outstripping supply. Delivering AI services closer to end users requires solving three constraints simultaneously: data centre space and power, low-latency global connectivity, and access to high-performance GPUs.

Leveraging its footprint of more than 1,100 data centres across 31 countries, together with Latitude.sh’s platform capabilities, Megaport intends to address these challenges through a decentralised approach. The Globally-Distributed AI Inference Cloud, of which the GPU Pool is a foundational component, will give enterprise customers access to GPUs alongside the CPU, memory, storage and networking infrastructure needed for production AI inference deployments.

Chief Executive Officer Michael Reid described AI inference as one of the biggest infrastructure opportunities of the coming decade, noting that the contracts reflect accelerating demand for globally-distributed inference infrastructure and that AI inference is a global infrastructure challenge rather than simply a GPU problem.

Four new AI infrastructure contracts

The four committed contracts carry a combined total contract value of approximately A$458.9 million (US$330.4 million), irrespective of usage, and are expected to commence in the first half of FY27. Approximate ARR of A$199.0 million (US$143.3 million) will be recognised incrementally as hardware is deployed and becomes operational, with the full run-rate contribution expected by the end of the first half of FY27.

The contracts carry a mix of terms, with around 55% of ARR and around 70% of total contract value generated by 36-month agreements. They require approximately A$369.5 million (US$266.0 million) of incremental capital expenditure, primarily for high-performance NVIDIA GPUs supported by network and storage components, with an anticipated payback of about 27 months. The customers are US-based technology providers running AI applications and inference workloads, supported by institutional shareholders, and comprise a mix of new and existing customers whose identities have not been disclosed for competitive reasons.

An on-demand GPU Pool

Central to the strategy is an on-demand GPU Pool supported by A$350.0 million (US$252.0 million) of investment, commercialised through consumption-based pricing. The platform will provide automated provisioning, billing and operations, enabling rapid access to infrastructure without traditional deployment lead times.

The GPU Pool is designed to serve three objectives: attracting new customers seeking immediate access to AI infrastructure, providing existing customers with flexible burst capacity, and creating a pathway to convert short-term usage into longer-term contracted deployments. Megaport expects it to take six to nine months to procure and deploy the servers, a further three to six months for them to ramp to optimal utilisation, and a 16-to-22-month payback target thereafter.

Trading update

Megaport reported continued growth across both divisions. For the April 2026 period, Network ARR rose 25% year-on-year on a constant currency basis to A$277.7 million, while net revenue retention by logo increased four percentage points to 113%.

In the Compute division, ARR excluding the strategic contracts announced between 27 April and the latest update grew 40% to A$83.9 million since the Latitude.sh acquisition. Including those strategic contracts, pro forma Compute ARR reached A$385.2 million, a 6.4-times increase since the acquisition. The strategic contracts contribute A$747.8 million of total contract value and A$301.3 million of ARR. On a group pro forma basis, total ARR is A$662.9 million.

Funding and the entitlement offer

To fund the capital expenditure, Megaport launched a fully underwritten 1-for-3.08 pro rata accelerated non-renounceable entitlement offer to raise approximately A$827.3 million. New shares will be issued at A$14.30 each, a 10.9% discount to the theoretical ex-rights price and a 13.9% discount to the last close of A$16.61 on 1 June 2026.

Proceeds will fund A$369.5 million of hardware to service the new contracts, A$350.0 million for the GPU Pool, A$19.3 million of transaction costs, and A$88.5 million for balance sheet capacity and future growth. The offer comprises institutional and retail components, with the retail offer open only to eligible shareholders in Australia and New Zealand. Chief Financial Officer Leticia Dorman said Megaport would remain in a strong net cash liquidity position following the raise, with pro forma liquidity as at 31 December 2025 of approximately A$287.6 million, and that debt is expected to become a more permanent part of the capital structure.

Guidance and outlook

FY26 revenue guidance has been tightened to between A$307 million and A$315 million, reflecting momentum in the Network business, while FY26 EBITDA margin guidance of 21% to 24% of revenue and Group capital expenditure guidance of A$90 million to A$100 million remain unchanged. The guidance does not reflect the impact of the strategic contracts. Over the medium term, Megaport expects continued investment in storage, network and CPUs to expand its distributed footprint.

Conclusion

The announcement marks a substantial step in Megaport’s transformation from a network connectivity provider into a combined network-and-compute platform targeting the fast-growing AI inference market. By pairing committed customer contracts with a flexible on-demand GPU Pool and a sizeable capital raising, the company is positioning itself to capture distributed inference demand, though the ultimate returns will depend on deployment timing, utilisation and execution across its global footprint.