Shares in Megaport Ltd (ASX:MP1) rocketed on Thursday after the network-as-a-service provider unveiled a trio of major AI infrastructure contracts worth a combined US$182.9 million (A$254 million), strengthening investor confidence in the company’s expanding push into high-performance compute.
The stock was up about 27% as of late Thursday afternoon, surging by as much as 37% earlier in the session.
The sharp move followed Megaport’s announcement that subsidiary Latitude.sh had secured three fixed-term contracts spanning GPU, CPU, networking and storage infrastructure for two US-based AI technology customers.
The agreements represent around US$65.2 million (A$90.6 million) in annualised recurring revenue once fully deployed, with about 90% of the total contract value tied to initial three-year terms.
AI infrastructure demand keeps accelerating
The announcement lands at a time when investors are increasingly rewarding companies exposed to the global AI infrastructure buildout, particularly businesses tied to data centres, cloud connectivity, compute deployment and networking capacity.
Megaport has traditionally been known for its software-defined connectivity platform, allowing enterprises to rapidly connect between cloud providers, data centres and networks. But the company’s acquisition of Latitude.sh last year signalled a broader ambition to move deeper into AI infrastructure and high-performance compute services.
The latest contracts appear to reinforce that strategy.
Megaport said the agreements covered GPU, compute, network and storage infrastructure supporting AI applications and inference workloads.
Importantly, the contracts are structured around committed long-term revenue regardless of customer usage levels, giving the company greater revenue visibility as it scales its compute offering.
One of the customers is already an existing Megaport client, which management framed as evidence of successful cross-selling between the company’s network and compute businesses.
Latitude acquisition begins to show strategic value
The market reaction also reflects growing investor attention on whether Megaport can successfully evolve from a pure connectivity provider into a broader AI infrastructure platform.
Since acquiring Latitude.sh, the company has been positioning itself as a provider of globally distributed, automated infrastructure capable of supporting the next wave of AI deployment — particularly inference workloads that increasingly require compute resources closer to end users.
“We are at the forefront of an accelerating inflection point across the industry,” said Megaport CEO Michael Reid.
“As use cases shift from AI foundation models to inference and the edge, Megaport is becoming an essential platform for powering the applications of tomorrow with globally distributed, automated infrastructure.”
That broader AI positioning has become increasingly important across global markets as investors look beyond the semiconductor giants themselves and towards the infrastructure layers supporting AI deployment.
Companies linked to networking, fibre, power, cooling, cloud orchestration and edge compute have all attracted renewed investor attention over the past year as hyperscalers race to expand capacity.
Big investment, but strong payback profile
The contracts will require roughly US$101 million (A$140.3 million) in incremental capital expenditure, largely tied to Nvidia Corp (NASDAQ:NVDA, XETRA:NVD) GPU hardware and associated infrastructure.
Megaport said the projects carried an estimated payback period of around two years and would ultimately leave the company with long-life compute assets that could later be redeployed across the broader Latitude.sh platform. The company plans to fund the investment through a combination of existing cash reserves and a newly upsized A$150 million debt facility.
Management also reaffirmed FY26 revenue and EBITDA guidance for the combined group.
For investors, the announcement appears to have reinforced a growing view that Megaport is becoming more directly leveraged to one of the market’s biggest structural themes: the global scramble to build the infrastructure needed to support AI at scale.