Papua New Guinea has stepped onto the sovereign compute map with the launch of Kumul Cloud Infinity, a new data center and cloud platform that Datec PNG built in partnership with Telikom PNG. The facility marks the country’s first AI-ready data center and signals a deliberate shift away from decades of dependence on offshore infrastructure. Government agencies, banks, universities, and private enterprises across the Pacific nation now have a domestic option for hosting cloud, data and AI workloads within Papua New Guinea, complementing existing domestic data center capacity.
The project arrives at a moment when sovereign AI and cloud infrastructure are emerging as active priorities across the Asia Pacific region, with Australia also advancing domestic AI infrastructure. Papua New Guinea’s entry into this race is modest in scale compared with larger economies, yet its symbolic weight is considerable for a nation charting its own digital independence. What Datec and Telikom have built is as much a statement of intent as it is a piece of physical infrastructure.
A New Digital Foundation for the Pacific
Datec and Telikom unveiled Kumul Cloud Infinity at a launch event that framed the facility as a turning point for the country’s technology sector. Datec, a longtime provider of ICT services in Papua New Guinea, positioned the data center as the anchor for a broader sovereign cloud strategy that touches government, banking, healthcare, and education. Telikom PNG, the state linked telecommunications operator, joined the project as a partner, lending its network reach to the new platform. Together, the two companies are betting that local businesses and public institutions will choose a domestic cloud over some of the international providers they have relied on for years. Datec has not disclosed pricing, service tiers, or a public roadmap for how quickly it expects the platform to draw workloads away from established overseas vendors.
Inside Kumul Cloud Infinity’s AI Engine
At the center of the new facility sits a deployment of Nvidia H200 GPUs, according to Datec chairman Moses Maladina, who spoke at the launch. The company has not disclosed how many GPUs it installed at the site, nor has it released details on the data center’s overall IT capacity or physical footprint. Maladina described the significance of the buildout directly: “We now have the capability here in PNG to run AI workloads, analyze large amounts of information, and support sophisticated AI applications, while at the same time providing cloud computing and data storage services required by Government, state-owned enterprises, and private businesses.”
That statement positions the facility as a dual-purpose asset, serving conventional cloud and data storage requirements alongside emerging demand for AI compute in a market where comparable sovereign AI capability remains limited. However, without published capacity figures, it remains difficult for outside observers to gauge how the site stacks up against data centers built for AI training or large scale inferencing elsewhere in the region.
Ending Decades of Offshore Data Dependency
The launch lands against a backdrop that Telikom Group CEO Amos Tepi laid out in stark terms in a recent LinkedIn post. Tepi wrote: “Since digital technology arrived in PNG, almost 100 percent of our data has been managed and stored abroad. For Government agencies, banks, hospitals, universities, businesses and citizens, this raised questions about location, governing laws, access and restoration during disruptions, and national control over strategically valuable data.” That reliance on foreign infrastructure has raised questions around where sensitive information is stored, which jurisdictions govern it, how organizations can access it, and how they can restore services during disruptions. Tepi framed Kumul Cloud Infinity as a direct answer to that vulnerability, writing: “Kumul Cloud Infinity provides a credible national response.
Hosting systems and data within PNG’s jurisdiction advances data residency and gives organizations greater control, confidence and visibility while supporting national resilience and innovation.” His comments place data residency, rather than raw compute power, at the heart of the project’s value proposition. For a country where Telikom’s CEO has said organizations have historically managed and stored almost all data abroad, even a new domestic AI and cloud facility represents a meaningful change in where organisations can host and process their data.
Building Toward Redundancy in Lae
Datec is not stopping at a single site. The company has confirmed it is developing a second data center in Lae, Papua New Guinea’s second-largest city, to provide geographic redundancy for the platform. Datec has provided no timeline for when that facility might open, and Datec has offered no further detail on its scope or specifications. Geographic redundancy matters in a country exposed to seismic activity and infrastructure disruptions, where a second site can strengthen resilience and reduce reliance on a single facility. A second site in Lae could extend Datec’s reach beyond the capital region, potentially opening the platform to businesses and institutions in other parts of the country. Until construction details emerge, the Lae project remains a stated ambition rather than a fully scoped buildout.
Before Kumul Cloud Infinity, Papua New Guinea’s data center footprint was thin by almost any regional measure. Datec’s own website does not detail its existing data center footprint, leaving outside observers with limited visibility into how much infrastructure the company already operates. That limited data center footprint helps put into context why so much of the nation’s data has historically been managed and stored abroad, as Tepi noted, even though domestic data center and cloud services already exist. Kumul Cloud Infinity does not just add a new facility to the map; it meaningfully expands the total pool of AI ready infrastructure available inside the country’s borders. The gap between Papua New Guinea and larger regional markets remains substantial, but the new facility adds dedicated sovereign AI capability to the country’s existing domestic data center infrastructure.
Australia’s Parallel Sovereign AI Push
Papua New Guinea’s move did not happen in isolation. The same week saw Australia advance its own sovereign AI ambitions, as AI infrastructure company SCX.ai announced a partnership with DataDirect Networks to build a sovereign AI inferencing cloud. The collaboration pairs SCX’s ASIC-based infrastructure with DDN’s Infinia data intelligence platform, aiming to give Australian enterprises, government agencies, and research institutions an inferencing cloud that keeps workloads inside the country’s borders. SCX already operates a sovereign AI node built on SambaNova processors at Equinix’s SY5 data center in Sydney, and that node will now see DDN’s Infinia platform integrated into its architecture.
What Sovereign Compute Means for the Pacific Region
Taken together, the Papua New Guinea and Australia announcements illustrate an emerging pattern in which countries are developing local capacity to keep critical data and AI workloads under greater national control. Smaller and mid-sized economies are no longer content to treat cloud and AI infrastructure as something imported wholesale from larger markets; rather, they are building local capacity to keep critical data and workloads under national control. Papua New Guinea’s approach leans on a single flagship facility backed by Nvidia H200 GPUs, while Australia’s strategy layers ASIC-based infrastructure with a specialized data platform aimed squarely at inferencing economics.
As a result, the region now has two distinct models for sovereign AI unfolding in parallel, one centered on foundational capacity building and the other on optimizing performance at scale. In turn, both projects raise the same unresolved question for observers watching from outside: whether announced capability will translate into published capacity, adoption numbers, and measurable performance once the initial launch enthusiasm fades. Right now, Kumul Cloud Infinity gives Papua New Guinea a new domestic foothold in sovereign AI and cloud infrastructure, adding capabilities that existing data centres did not previously provide at a comparable level.


