Megaport Secures New AI Contracts, Announces Capital Raise During Halt: The Latest

Megaport Limited (ASX: MP1) will emerge from a trading halt on Thursdau with a transformative announcement that repositions the network-as-a-service provider as a significant player in AI infrastructure, securing four new contracts worth A$458.9 million and unveiling plans for a global GPU cloud supported by a fully underwritten A$827.3 million equity raise.

Megaport shares had surged 84.35% over the past month of trading, trading at A$16.61 before entering a trading halt on 2 June. The company’s year-to-date performance prior to the halt showed gains of 34.82%, reflecting growing market enthusiasm for its pivot into AI infrastructure. The entitlement offer has been priced at A$14.30 per new share, representing a discount to pre-halt trading levels and establishing a new reference point that markets will likely gravitate toward when trading resumes.

The pricing reflects a delicate balance: high enough to demonstrate confidence in the AI infrastructure opportunity, yet discounted sufficiently to ensure strong institutional support for what represents one of the largest capital raises in the company’s history. The fully underwritten nature of the offer eliminates funding risk but introduces material dilution for existing shareholders who do not participate.

AI Infrastructure Contracts

The four new AI infrastructure contracts carry a combined Total Contract Value of approximately A$458.9 million and will require around A$369.5 million in capital expenditure, predominantly for NVIDIA GPUs alongside network and storage infrastructure. The contracts are with US-based technology providers running AI applications and are expected to commence in the first half of 2027.

This announcement follows earlier momentum in May 2026, when Megaport’s Latitude.sh subsidiary secured three binding fixed-term contracts worth US$182.9 million (approximately A$254 million) with two US AI technology companies. Together, these deals represent over A$710 million in contracted AI infrastructure revenue, marking a step-change in scale for a company that has traditionally focused on software-defined networking.

The contracts demonstrate a structural shift in Megaport’s business model. On a pro forma basis, the Compute division’s Annual Recurring Revenue has jumped to A$385.2 million, now representing the majority of total Group ARR of A$662.9 million. Meanwhile, the legacy Network business continues to show strength, with Network ARR increasing 25% year-on-year to A$277.7 million and Net Revenue Retention holding at 113%, indicating robust upsell and expansion within the existing customer base.

Global GPU Pool Strategy

Beyond the contracted deals, Megaport is committing approximately A$350 million to create a new global GPU Pool, designed to provide enterprise customers with flexible, on-demand access to AI infrastructure. The initiative addresses surging demand as workloads shift toward latency-sensitive AI inference applications, where proximity to data sources and end-users becomes critical.

The company plans to deploy its Globally-Distributed AI Inference Cloud across more than 1,100 connected data centres spanning 31 countries, leveraging Megaport’s existing network footprint. The GPU Pool infrastructure is scheduled for rollout over the next 6-9 months, with servers expected to reach optimal utilisation within 3-6 months after deployment. This timeline means the economic benefits will be weighted toward late FY26, FY27, and beyond, creating a significant lag between capital deployment and revenue realisation.

The strategic rationale centres on differentiation: by combining a global, software-defined network fabric with on-demand compute and storage, Megaport aims to enable low-latency inference positioned close to data and users. This approach targets a market gap between hyperscaler cloud offerings and specialised, single-location GPU providers, offering enterprises the flexibility to deploy AI workloads across distributed locations without the complexity of managing multiple vendor relationships.

Capital Raise

The A$827.3 million entitlement offer is fully underwritten, sharply reducing execution risk for what represents an exceptionally capital-intensive growth plan. Eligible shareholders as of 5 June can participate according to their allocation, with both institutional and retail components included in the structure.

The raise comes on top of previous capital injections during late 2025 and early 2026, when Megaport raised approximately A$200 million through institutional placements and share purchase plans to fund the acquisition of Latitude.sh for around A$459 million. That acquisition provided the compute and GPU infrastructure capabilities that enabled the current contract wins and form the foundation of the GPU Pool strategy.

For existing shareholders, the cumulative dilution represents a significant reset of the equity story. Those who participated in earlier raises have already experienced one dilution event, and the current A$827.3 million offer, substantially larger than previous rounds, will materially increase the share count. The capital structure is being reweighted toward funding aggressive growth in AI infrastructure, a marked departure from the asset-light network model that characterised Megaport’s earlier years.

Guidance

Megaport has tightened its FY26 revenue guidance to a range of A$307 million to A$315 million, while maintaining unchanged EBITDA margin and capital expenditure guidance for the existing business. The unchanged operating metrics suggest management confidence in maintaining profitability and operational leverage within the core network business, even as substantial capital is deployed into the new GPU infrastructure.

The company plans to provide further updates on progress at its full-year results announcement scheduled for August 2026, which will be a critical checkpoint for markets assessing execution against the ambitious deployment timeline. Key performance indicators to watch include GPU utilisation rates, additional contract signings beyond the initial four deals, and the pace at which the GPU Pool attracts enterprise tenants beyond the contracted baseline.

The pivot from network automation to operating a global, capital-intensive GPU fleet introduces materially higher operational complexity and competitive risk. Megaport will be competing against hyperscalers with vastly greater resources, specialised AI infrastructure providers, regional data-centre operators, and an expanding field of sovereign and private cloud initiatives, all deploying NVIDIA GPU-based inference capabilities.

The timing mismatch between capital deployment and revenue realisation adds near-term pressure. The A$369.5 million in contracted capex will be spent over the next 12-18 months, while the four contracts only commence in the first half of 2027. If AI demand moderates or customer-specific ramps underperform expectations, markets face a period of depressed free cash flow before the revenue catch-up materialises.

What Next for MP1

The transformation of Megaport from a pure network-as-a-service provider into an integrated AI infrastructure platform represents a fundamental repositioning of the business. Management is betting that the combination of global network reach, software-defined automation, and distributed GPU capacity will create a differentiated offering for enterprises navigating the complexity of deploying AI workloads at scale.

The integration of Latitude.sh provided the initial compute capabilities and customer relationships that enabled the contract wins. The current capital raise and GPU Pool initiative represent the next phase; building out a scaled, globally distributed AI inference cloud that can serve both contracted and on-demand workloads. If successful, the strategy could position Megaport as a meaningful alternative to hyperscaler AI infrastructure, capturing a share of enterprise spending as workloads diversify beyond centralised cloud environments.

The risk is that execution proves more challenging than anticipated, competitive intensity erodes pricing and margins, or the AI spending cycle moderates before Megaport achieves scale and profitability in its compute division. The company is now firmly tethered to the AI infrastructure cycle, with valuation and risk profile tracking AI demand dynamics rather than the more stable connectivity market that previously defined the business.

With Megaport shares set to resume trading tomorrow, it is likely to be a volatile day, with pull towards the raise price, and some push on the deal news. Not a week for the faint of heart.