Galasso says New Zealand's distance is now its advantage

About Datagrid

Datagrid is developing the Makarewa campus north of Invercargill, the largest data centre project by disclosed capacity in New Zealand's build. The site holds a 360MW grid connection and an initial data centre design of 280MW, consented in March 2026 by Southland District Council, Environment Southland and Invercargill City Council on a non-notified basis, which RNZ and 1News reported as the first New Zealand data centre approved that way. RNZ has reported construction due to start within a month of 27 July 2026, more than 1,200 construction jobs, 50 to 80 permanent roles once the campus is running, and an opening target of 2028.

The company is building the cable as well as the campus. Datagrid contracted phase one of the Tasman Ring Network with ASN and OMS Group on 31 July 2026, a 2,300km system from Invercargill to a landing at Torquay in Victoria, specified at a minimum of eight fibre pairs and at least 160Tbps of design capacity, with service targeted for 2029. Mercury became a shareholder in July 2026, paying US$30 million for 12.7% four months after signing a 140MW, 15 year power purchase option over half the campus's initial design.

Rémi Galasso founded Hawaiki, the 15,000km trans-Pacific cable linking Australia, New Zealand, American Samoa, Hawaii and the US west coast that went into service in 2018, and later founded Intelia, whose Te Waipounamu design the Tasman Ring superseded. He is Datagrid's founder and chief executive, and he put the following to Datacentres NZ on how far New Zealand's case has come and what still has to be built.

Q&A with Rémi Galasso, Datagrid's founder and chief executive

What follows is Galasso's written contribution to Datacentres NZ, published as submitted. His answers appear in quotes; our independent analysis follows where useful.

Q: You've long made the case that New Zealand could become a cloud and data haven: massive renewable power, cool climate, political stability. How much of that has been realised, and how big can this get?

"The vision is turning into reality, and faster than most expected. The world is running short of compute for the AI build-out, and geopolitical uncertainty has pushed the hyperscalers and AI companies to rethink where they put it. What used to be seen as New Zealand's disadvantage, our distance, our remoteness, is now clearly an advantage: stable, secure, over 85% renewable, naturally cooled and far away from conflict zones. We're not claiming to rival the mega-hubs on raw scale. But as a trusted place to train and store the world's most valuable data, New Zealand can be one of the best locations on the planet, and this is only the beginning."

Q: You've said 'the Americans are coming, and they are coming big.' When hyperscalers and AI companies sit across the table from you and ask about New Zealand, what's on their checklist, and what are NZ's advantages in those conversations?

"The American AI companies are leading the race, on the chips and on the applications, and they're moving fast into our region. Australia is a natural first target and has a lot to offer, I'd never argue otherwise. New Zealand competes on a different scale, but on the fundamentals that increasingly decide these deals, we're hard to beat: power efficiency, water efficiency, and genuinely green electricity. On PUE, WUE and carbon, Southland can outperform anything on offer across the Tasman. When those advantages sit next to political stability and security, the conversation gets very interesting."

Q: You've reframed this region from the dead end of the internet to the middle of a US-to-Middle East route. Walk us through NZ's connectivity map in 2030: where do the cables land, and how do compute and connectivity sequence together?

"For years the assumption was that this region sat at the dead end of the internet. That's changing. The South China Sea has become difficult for the hyperscalers, hard to get permits to lay new cable, hard even to repair what's there, so the industry is actively building alternatives. One of the most compelling runs from the US west coast down through the Tasman Sea, south of Australia, and on to Singapore, India and the Middle East. We call it the Great Southern Route, and it puts Sydney and Melbourne squarely in the middle of it. Our neighbours will make the most of that, and so can we, which is exactly why we're building the Tasman Ring Network, a 2,300km system that will connect Invercargill to Melbourne. Compute and connectivity have to sequence together: the Southland campus is the anchor tenant that makes the cable viable, and the cable will unlock the potential of the South Island for digital infrastructure, IT companies, Universities and Research Center. Invercargill will become a hub, and the Great Southern Route might well be extended up to Chile at some stage."

Analysis: Phase one targets service in 2029, a year after the campus's 2028 opening target, and is the only leg under contract, with an option to build the rest later. Galasso put the latency cut between the South Island and Australia at up to 45%, and Datagrid head of connectivity Pierre Goiffon at a minimum of 19 milliseconds round trip.

Q: You told us the 360MW is grid capacity and the final data centre design is still being optimised, and that Datagrid will achieve the best WUE in the region. How does the design actually get there, and how much of that is the climate versus the engineering?

"Our engineers are still optimising the design, and a lot of it depends on the chips our future tenant will choose, some generations need cooler data halls than others. In the best case, Southland's climate lets us run a closed-loop liquid system cooled naturally by the outside air, which means that for most of the year we don't touch water at all. I know water use is a real concern for people in New Zealand, and I want to address it head-on: this facility is being engineered from the ground up to target one of the lowest water consumption in Asia Pacific, and for much of the year it will draw effectively none. So it's both: the climate gives us the opportunity, and the engineering is what actually captures it. We expect to finalise the design in the coming weeks and put firm water numbers on the table."

Q: Given the transformational scale of what's being built in New Zealand, and the global competition for talent, how can the country upskill its workforce?

"I'm not worried about our construction sector's ability to build and deliver a facility of this scale on time, that capability is here. The real challenge, and we share it with Australia, is the specialist skills an AI data centre needs, both technicians and engineers. But New Zealand has world-class universities and research centres on both islands, and I'm confident they'll adapt quickly to demand and stand up genuinely state-of-the-art programmes, for local students and to attract international talent as well. A project like this is precisely the kind of anchor that pulls that ecosystem into being."

Q: Mercury didn't just sell you power, it bought into the company. Is the gentailer-as-shareholder model the template for how New Zealand finances digital infrastructure, and do you expect others to follow?

"This is a model that's increasingly common in the US, and we're the first to bring it to New Zealand. Whether others follow, we'll see, but it makes a lot of sense here. Importantly, partnering with Mercury doesn't lock us in: we can and will continue to work with other gentailers, and we have proper information barriers in place to protect confidentiality and keep competition between generators fair. For us it's not just a supply contract, having a leading renewable generator invested alongside us is a real vote of confidence in the project."

Makarewa in figures, and where each comes from

Public reporting has carried several cost estimates for Makarewa, and capacity is often quoted against the grid connection when the data centre design is the smaller number.

Source: Datagrid, Mercury, Environment Southland and the Tasman Ring contract announcement, with Galasso's written responses dated 3 August 2026.

What this means for operators and investors

Galasso is the developer, so every forward figure here is his to prove, and the table above exists to keep his claims and the record apart. The water commitment is the one worth diarising: Environment Southland's 220 million litre ceiling has been the only number on the public record since March, and Galasso has now attached a timeframe to a second one.

Galasso calls the Southland campus the Tasman Ring’s anchor tenant, reversing the usual account in which a cable exists to serve a campus. On current targets the cable arrives a year after the campus opens, so the first traffic out of Makarewa runs on the routes New Zealand already has.

Galasso names the specialist technician and engineer shortage as one New Zealand and Australia share, and Australian demand for those roles is tracked in data centre jobs in Australia.

Mercury has taken both a shareholding and a board seat in a developer it has optioned power to, a structure Galasso says is common in the United States and new to New Zealand. A second gentailer doing the same would show the model has carried.

Key takeaways

• In an exclusive Q&A, Datagrid founder and chief executive Rémi Galasso says New Zealand’s remoteness has become an advantage, and that the shift has come faster than most expected.

• Galasso says Southland can outperform Australian sites on power efficiency, water efficiency and carbon, figures Datagrid still holds internally for Makarewa.

• Galasso says Datagrid will finalise the campus design within weeks and publish firm water volumes alongside it, against a consent ceiling of about 220 million litres a year.

• Tasman Ring phase one is contracted to run 2,300km to a Torquay landing, with a 2029 service target, a year after the campus’s 2028 opening target.

• Galasso says Datagrid remains free to contract with other gentailers alongside Mercury’s 12.7% shareholding, with information barriers in place to keep competition between generators fair.