Megaport (ASX:MP1) has spent the past decade building its name on network-as-a-service — the plumbing that lets enterprises connect securely and on Demand to the world’s data centres and clouds. In a matter of weeks this year, that identity was significantly broadened. The Brisbane-based company has signed four AI infrastructure contracts worth roughly $458.9 million, committed to hundreds of millions in Nvidia GPU capex, and raised $827.3 million in a fully underwritten entitlement offer to fund the expansion.

The scale of the funding and contract wins has put Megaport firmly in focus as it moves from a connectivity operator toward an AI-infrastructure provider. The size of the raise — one of the larger secondary raisings on the ASX technology boards this year — signals a company betting its Balance Sheet on the thesis that distributed, on-demand compute could become as valuable as distributed connectivity once was.

Whether that strategy pays off, and on what timeline, is now the central question for investors watching MP1 heading into full-year results.

What Happened?

On 3 June 2026, Megaport emerged from a Trading halt to announce four new AI infrastructure contracts with combined total contract value of approximately $458.9 million. Delivering on them requires an estimated $369.5 million in incremental capital expenditure, primarily for Nvidia GPUs, networking and storage hardware.

Alongside the announcement, the company unveiled an $827.3 million capital raise structured as a 1-for-3.08 accelerated non-renounceable entitlement offer priced at $14.30 per share, including roughly $350 million earmarked for a globally distributed, on-demand GPU pool.

The institutional leg closed on 5 June 2026, raising about $518 million with take-up near 99% — a result CEO Michael Reid described as strong support from institutional shareholders. The retail offer followed, opening 11 June and targeting roughly $309 million, completing oversubscribed with new shares trading from mid-June.

Megaport hands down full-year FY26 results on 20 August 2026, providing an important test of how the AI contracts and associated spending are progressing against guidance.

Why It Matters

The shift matters because it changes Megaport’s Earnings mix and risk profile together. Historically, Revenue has come from relatively capital-light, high-margin network connectivity. The AI contracts pull the company into a materially more capital-intensive Business — buying and operating GPU hardware — with different economics, including Depreciation and utilisation risk.

Management has been explicit that the roughly $199 million of annual Recurring Revenue expected from the AI contracts will only be recognised incrementally as hardware is deployed, with deployment timelines of six to nine months and further ramp-up of three to six months after that.

The revenue opportunity is substantial, but back-ended. Capital Expenditure occurs well before the majority of AI-related revenue is recognised, making execution and utilisation critical to the Economics of the strategy.

The Bigger Story

Megaport’s pivot sits inside a larger industry story: the scramble by cloud, telecom and infrastructure companies to capture AI inference demand as enterprises move from Training large models to running them in production, often close to end users and existing data flows.

Unlike hyperscale training clusters concentrated in mega-campuses, inference workloads are expected to be more geographically dispersed — a dynamic that plays to Megaport’s existing strength in distributed, multi-market connectivity.

The logic is that owning both the network fabric and the compute layer lets Megaport offer a differentiated, low-latency AI service rather than competing purely on price against hyperscalers.

It is a crowded field, however, with well-capitalised cloud providers, colocation giants and specialist GPU-cloud start-ups chasing the same opportunity. Nvidia’s own Supply and pricing power also remain important variables for any company building a hardware-heavy compute business.

Bull Case

• Contracted AI revenue of roughly $199 million in ARR, once hardware is deployed, would materially lift scale and diversify Megaport beyond network connectivity.
• The institutional offer’s ~99% take-up rate signals strong Shareholder confidence in the strategy and management’s execution credibility.
• The core network business remains healthy, with 25% ARR growth and 113% net revenue retention, providing a stable base while AI scales.
• A fully funded balance sheet reduces near-term financing risk for the capital-intensive GPU rollout.

Bear Case / Risks

• The AI contracts pivot Megaport into a far more capital-intensive, hardware-dependent model with utilisation, depreciation and obsolescence risks absent from its legacy business.
• Revenue recognition is back-ended, with six-to-nine-month deployment windows and further ramp-up, so AI revenue may not show up materially until FY27.
• Nvidia GPU supply, pricing and refresh cycles sit outside Megaport’s control and could pressure margins or delay deployment.
• Significant dilution from the $827.3 million raise raises the bar for per-share value creation from the new segment.

What Investors Should Watch Next

• FY26 full-year results, due 20 August 2026, for updated guidance and commentary on AI deployment progress.
• Milestones on GPU delivery and data-centre commissioning against the six-to-nine-month timeline.
• Early revenue recognition from AI contracts as hardware becomes operational through FY27.
• Ongoing core Network ARR growth and net revenue retention as a gauge of business health.
• Any further AI contract wins or competitive responses from rival cloud and colocation providers.

Conclusion

Megaport has made a decisive, well-funded bet that distributed AI inference infrastructure will become as essential as the connectivity it built its name on. The $827.3 million raise, near-full institutional take-up and oversubscribed retail offer provide substantial funding for the strategy, while the underlying network business continues to grow.

But the AI contracts also convert Megaport into a more capital-intensive, execution-dependent story, with significant spending required before the majority of the associated revenue is recognised. Hardware costs, deployment timing, utilisation, competition and GPU supply will all determine whether the expansion delivers attractive returns.

The 20 August results will provide an early test of whether Megaport's AI infrastructure strategy is beginning to translate its contracted opportunity into delivered operating performance.