Amazon agreed to buy backup generators from Generac, the company best known for the generators that keep homes running during blackouts. These machines do the same job for the giant computer buildings behind online services, starting up when the regular power fails.
About $2.4 billion worth is expected to be delivered in 2027 and 2028. Amazon also gets the right to buy Generac stock at a fixed price, and that right grows the more generators Amazon buys, up to $8 billion.
The first $2.4 billion has a date on it. The rest only happens if Amazon keeps building.
On September 16, 2026, Generac Holdings (NYSE: GNRC) and Amazon.com signed a long-term supply agreement for backup power generators for Amazon data centers, according to an 8-K Generac filed the same day. Initial deliveries are expected to total $2.4 billion across 2027 and 2028. Alongside the supply deal, Generac issued an Amazon subsidiary a warrant to buy up to 1,693,745 Generac shares at $200.9266 each. Some of those shares vested immediately, and the rest vest as Amazon’s generator payments accumulate, up to a total of $8 billion.
Yahoo Finance reported that Generac shares rallied on the news.
Executive Summary
Generac’s filing makes Amazon a named buyer of data-center backup power, the generators that carry a facility through a grid outage once its batteries run down. It puts two numbers on the relationship. The first is $2.4 billion in expected deliveries over two years. The second is an $8 billion ceiling on the purchases that can unlock Amazon’s equity stake.
Those two figures measure different things. The $2.4 billion is a delivery expectation with dates attached. The $8 billion is the spending level at which Amazon’s warrant fully vests, and the warrant runs until September 2033. Most of the headline figure is therefore contingent: it turns into revenue only if Amazon keeps building data centers that need Generac equipment.
The structure also shows who is paying for what. Generac is giving its customer equity upside in return for volume. For Generac shareholders, that is a trade of some future dilution for a large anchor buyer. For the wider data-center supply chain, it signals that backup generation is now valuable enough to be locked in years ahead with financial ties that go beyond a purchase order.
The $8 Billion Is a Ceiling, Not an Order
The 8-K describes two separate instruments. The long-term supply agreement governs what Amazon buys, and the only scheduled volume it discloses is the $2.4 billion in initial deliveries expected in 2027 and 2028. The warrant governs what Amazon earns, and it vests in tranches “contingent upon aggregate gross payments, net of certain offsets” from Amazon for data-center backup generators, “up to a total of $8 billion.” The $8 billion is the point where the incentive stops growing. It is not a commitment to spend that much.
Scheduled deliveries cover 30% of that ceiling. The remaining $5.6 billion has no timetable in the filing, only an outer boundary: the warrant can be exercised on or before September 16, 2033. Every new hyperscale data center needs its own backup generation, sized to its electrical load. Whether Generac’s revenue approaches the $8 billion mark therefore depends on how many facilities Amazon builds over the next seven years, how large they are, and how much of that generator demand Generac wins against other suppliers.
This matters to three groups. Investors should model $2.4 billion as the disclosed baseline and treat anything above it as dependent on Amazon’s construction pace. Generac’s factory planning has to support a two-year delivery ramp without knowing what comes after 2028. Other generator suppliers now face a buyer that has a financial incentive to direct its volume toward one vendor.
Generac Is Paying for Volume in Equity
A warrant is the right to buy shares later at a fixed price. Here the supplier grants it to the customer. That makes it a volume incentive paid in stock rather than cash. Of the 1,693,745 warrant shares, 307,954 (about 18%) vested at signing, which works as an upfront signing incentive. The other 1,385,791 unlock only as Amazon’s payments grow. Spread evenly across $8 billion, that would be roughly one share for every $5,770 of generator purchases. The filing says only that vesting occurs in “multiple tranches” and does not give the actual schedule.
The economics favor Amazon when Generac’s stock trades above the $200.9266 strike price. Buying all 1,693,745 shares would cost about $340 million at that price. However, Amazon can choose a cashless exercise, receiving the net value in shares without paying cash. Generac has also granted registration rights, which make it easier for Amazon to sell those shares later. The result ties part of Amazon’s return to Generac’s share price, and that price should rise as Amazon buys more generators. Existing Generac shareholders bear the cost through potential dilution. The warrant’s strike price and share count are also subject to anti-dilution adjustments.
A Home-Standby Brand Takes a Hyperscale Test
Generac built its name on home standby generators. Data-center backup is a different business. The units are larger, reliability requirements are strict, and a hyperscaler’s construction schedule does not move because a supplier fell behind. Having Amazon as a named, contracted buyer is the clearest external validation of Generac’s data-center push so far. It also exposes Generac to a direct test: delivering $2.4 billion of equipment within two calendar years, to a customer that can measure its performance precisely.
Amazon’s broader funding shows how much money is flowing into these projects. Two days before the Generac agreement, Amazon’s own 8-K reported that it closed a £4.25 billion sterling bond sale, with coupons ranging from 5.200% on notes due 2029 to 6.650% on notes due 2045. Nothing in that filing links the proceeds to Generac or to data centers in particular. What it does show is a buyer raising capital at a scale where multi-year equipment commitments are routine. For Amazon, a warrant that pays off only if Generac performs is an inexpensive way to secure a supply chain it will need regardless.
Background
Generac Holdings, which trades as GNRC, is best known for home standby generators, the units that switch on automatically when a house loses power. Data-center backup power is a larger, more demanding version of the same product. Each facility needs enough on-site generation to carry its full electrical load through a grid outage, and hyperscale operators such as Amazon buy these systems in volume as they build new campuses.
Warrants tied to a customer’s purchases are an established way for suppliers to secure large buyers. The supplier grants equity upside that grows as the customer spends. Generac’s agreement follows that model, pairing a long-term supply contract with a warrant that vests against Amazon’s generator payments.
A factory near Milwaukee wants up to 300 new workers. It builds giant cooling systems for data centers, the warehouse-sized buildings full of computers that run websites and artificial intelligence tools.
Pay starts at $23 an hour, about $47,800 a year full-time. No factory experience is needed, and training happens on the job.
The surprising part is that the building opened a year ago. What the company needs now is people to run a second shift. Every one of those computer buildings needs cooling before it can switch on, so this hiring helps decide how fast they arrive.
Modine (NYSE: MOD), the Racine, Wisconsin-based thermal management company, said on September 23, 2026 that it is hiring for up to 300 long-term manufacturing positions at its data center cooling plant in Franklin, Wisconsin. The roles span first and second shifts. Production assembler pay starts at $23 an hour, second shift earns an extra $1 an hour, and benefits begin on day one. A walk-in hiring fair runs September 24 from 4 to 7 p.m., and a billboard campaign across the Milwaukee area will run while hiring continues over the coming months.
The 153,000-square-foot Franklin plant opened in September 2025 and makes Modine’s Airedale data center cooling systems. The announcement comes a week before Modine completes the spin-off of its Performance Technologies business into a combination with Gentherm, which is expected to close October 1.
Executive Summary
Modine’s announcement is a hiring notice, but it also shows where the company is in its data center cooling expansion. In July 2025 Modine announced added U.S. capacity for Airedale cooling products across plants in Franklin, Wisconsin; Grand Prairie, Texas; Grenada, Mississippi; and Jefferson City, Missouri. The Franklin building has now been operating for a year. Modine says the next step is people: “We’re now at a point where we need a larger, committed workforce,” said Karen Sutherland, Vice President of Human Resources for Data Centers, Americas.
That matters to anyone who depends on cooling equipment arriving on schedule. Once a plant exists, its output is set mainly by how many shifts it can staff and how quickly new workers become productive. Modine is recruiting mostly people without manufacturing experience and training them on the job. The speed of that process will shape how quickly the Franklin plant converts demand into shipped units.
The timing adds a corporate dimension. Filings made September 17 show Modine separating its Performance Technologies business on October 1. The company that remains, which Modine has said will be renamed Modexus Solutions, will be smaller and more concentrated on businesses like Airedale.
The Building Is Done. The Shifts Are Not.
A factory’s output comes down to three things: floor space, the hours that space is in use, and the productivity of the people working in it. Modine settled the first in September 2025, when it opened the 153,000-square-foot Franklin plant dedicated to Airedale data center cooling systems. This week’s announcement concerns the other two. Modine says the up-to-300 new hires will “help ramp up production to meet increasing demand,” and it is hiring across both first and second shifts, paying a $1-an-hour premium to fill the later one.
A second shift is usually the fastest and cheapest way to raise output from an existing plant. It needs no new permits, no construction, and no new utility connection. It needs people, and in Modine’s own words that is the requirement it has now reached. For the Franklin plant, the pace of growth now depends mainly on recruiting and training. Construction timelines matter less at this stage. This is the plain meaning of the title. Floor space in Franklin is fixed at 153,000 square feet, and extra output from it has to come from staffing more hours.
This affects data center developers and operators who are waiting on cooling equipment. A cooling supplier’s plant footprint is only a rough proxy for its delivery capacity. Shift coverage and workforce experience are closer measures. For Modine’s investors, the hiring pace becomes a leading indicator of whether Franklin can turn demand into revenue at the rate the company expects. Modine has not published any output targets for the plant.
What $23 an Hour and No Experience Required Signal
At $23 an hour and 40-hour weeks, a starting assembler earns about $47,800 a year before overtime, or about $49,900 on second shift. If all 300 positions were filled at the starting rate on full-time hours, base wages alone would come to roughly $14.4 million a year, before benefits and shift premiums. This is an illustrative calculation from the release’s figures, not a number Modine disclosed. Still, it shows that the ramp represents a meaningful recurring cost commitment.
The recruiting approach says more than the wage does. Modine says most entry-level roles need no prior manufacturing experience and that it provides on-the-job training. That widens the pool of applicants, but it also means new hires start at the bottom of a learning curve. Output per worker is likely to rise over months, not immediately. A walk-in fair with on-the-spot interviews and a Milwaukee-area billboard campaign point to hiring at volume. Modine also describes the roles as long-term. Hiring a permanent workforce rather than temporary labor suggests Modine expects demand to last, although that is the company’s own characterization and it has not tied the roles to a disclosed order book.
A Smaller Modine, More Concentrated on Cooling
The hiring announcement arrives days before a major change in Modine’s structure. According to an 8-K and the joint Gentherm–Modine release filed September 17, Modine will spin off its Performance Technologies business and merge it into Gentherm in a Reverse Morris Trust. In this structure, a unit is distributed to shareholders and immediately merges with another company, generally so the separation can be tax-free. The record date is September 28, and the distribution and merger are both expected October 1, 2026. Former SpinCo holders are expected to own about 43.6% of the combined company.
The filings also revise the cash side of the deal. Because falling overlap between Modine and Gentherm shareholders will increase the exchange ratio, Gentherm expects to issue about 2,902,466 additional shares. To offset that, the cash distribution to Modine before the merger is expected to fall from $210 million to $159 million, and Gentherm will pay its own shareholders a special dividend of about $58.35 million. The companies describe the adjustment as economically neutral, and final figures will be set at closing.
The Airedale data center cooling business stays with Modine. After October 1 it becomes a larger share of a smaller company, which Modine has said will be renamed Modexus Solutions. That raises the importance of execution at plants like Franklin. The analyst and investor day scheduled for November 18, 2026 is the next natural occasion for the company to put numbers on capacity, staffing and demand.
Background
Modine has worked in thermal management for more than 100 years. It is headquartered in Racine, Wisconsin, employs more than 11,000 people, and operates in North America, South America, Europe and Asia. It has reported two segments, Climate Solutions and Performance Technologies. Performance Technologies is being separated and combined with Gentherm on October 1, 2026, and the remaining company is due to be renamed Modexus Solutions. Modine’s data center cooling products are sold under the Airedale brand.
Data center cooling is the equipment that removes heat from server halls: chillers, air handlers and related systems that keep computing hardware within safe operating temperatures. As U.S. data center construction has grown, suppliers have added manufacturing capacity. Modine’s July 2025 expansion across four U.S. plants, including the Franklin facility that opened in September 2025, is its part of that buildout.
Google has been ordered to pay $463 million for breaking a European Union rule covering location data — the record of where a phone has been.
The sum is large in absolute terms, but small next to what a company of Alphabet’s size earns in a single three-month period. Its real weight is as a public price tag on handling personal data badly.
It matters beyond maps. The same European rules cover the data that artificial intelligence services quietly collect while they run. Any business using those services in Europe carries part of that risk.
AP News reported that Google has been fined $463 million for breaching a European Union rule governing location data — the positional records generated by phones and the apps running on them. The penalty lands on Alphabet’s core consumer business, where location signals underpin mapping, local search and ad targeting.
The amount places the decision among the more consequential European data-protection actions taken against a single company, and it arrives at a moment when every large cloud and artificial-intelligence provider is expanding the volume of behavioural data it collects, moves and stores inside the bloc.
Executive Summary
A $463 million penalty against Google for mishandling location data is, on its face, a consumer-privacy story. Read as an infrastructure story, it is something more useful: a published, enforceable price for a category of failure that until now most organisations have carried on their risk registers as an unquantified exposure.
Location data is a well-understood regulatory target. It is personal, it is continuous, it is generated as a by-product of a service the user came for, and it is retained by default. Those four properties describe location history. They also describe the telemetry produced by modern artificial-intelligence systems: prompt logs, retrieval traces, session identifiers, device signals, model-feedback records. None of that was at issue in this case. All of it sits under the same European legal regime.
For buyers of cloud and AI capacity in Europe, the practical consequence is that data-handling design — where logs land, how long they persist, who can read them, whether they can be deleted on request — moves from a compliance checkbox to a line item with a comparable number attached to it.
A Half-Billion-Dollar Number Is a Benchmark, Not Just a Bill
Regulatory fines do two jobs. The first is punitive and specific to the company. The second, and the more durable one, is that they establish a reference price. Before an enforcement action of this size, a compliance team arguing for budget to rebuild a retention pipeline is making an abstract case about tail risk. Afterwards, there is a figure in the room.
$463 million is not existential for Alphabet — it is a modest fraction of what the company earns in a quarter, and markets generally treat penalties of this scale as an operating cost rather than a structural event. But that is precisely why the number travels. It is small enough that Google can absorb it and large enough that a mid-sized enterprise, a cloud reseller or a data-centre operator offering managed services cannot. European data-protection law does not scale its rules down for smaller defendants; it scales the fine down, while leaving the obligation identical. The deterrent lands hardest on organisations with the least capacity to engineer around it.
Location Data Was the Test Case. Machine Telemetry Is the Larger Surface.
Nothing in this case concerns artificial intelligence, and it would be wrong to read it as an AI ruling. The mechanism it enforces, however, is agnostic about which product generated the data. European data-protection rules attach to personal data by its nature and its processing, not by the business line that produced it. What made location history legally fragile was that it was collected continuously as a by-product of another service, retained by default, and hard for a user to see, correct or erase.
Every one of those characteristics is present in the exhaust of a production AI system. Inference endpoints log prompts and completions for quality monitoring and abuse detection. Retrieval-augmented systems log which internal documents were fetched for which user. Agentic tooling logs actions taken on a user’s behalf, often against third-party systems. Fine-tuning and evaluation pipelines retain samples specifically because they are useful later — the retention is the point. When those records identify a person, directly or by combination, they are personal data under the same regime that produced this fine, and the same questions apply: what was the lawful basis, how long is it kept, can it be deleted on request, and where physically does it sit.
Who this affects is concrete. It affects enterprises signing AI contracts in Europe, who now have a sized precedent to point at when negotiating logging and retention terms. It affects model and cloud providers, whose default telemetry settings — often designed for debugging convenience and for model improvement — become contractual battlegrounds rather than footnotes. And it affects operators selling European capacity, because the market’s willingness to pay a premium for in-region processing is a function of how expensive the alternative looks. A published $463 million figure makes the alternative look more expensive than it did the week before.
The Infrastructure Response Is Boring, Expensive and Already Underway
The engineering answer to this class of risk is unglamorous: keep less, keep it in fewer places, and be able to prove both. In practice that means in-region inference so prompts never leave the jurisdiction, log pipelines with enforced time-to-live rather than indefinite retention, tokenisation or pseudonymisation of identifiers before they reach an analytics store, key management held by the customer rather than the provider, and deletion tooling that can actually reach into backups and derived datasets. Each of those adds cost — more regional footprint, more duplicated storage tiers, more egress between zones, more engineering time spent on data lineage rather than features.
That cost is the opportunity. Sovereign and in-region cloud offerings, European colocation with clear jurisdictional boundaries, confidential-computing hardware, and the data-governance tooling layer all sell against precisely this risk, and they have historically struggled to price it because the counterfactual was invisible. Enforcement makes the counterfactual visible. The likely near-term effect is not a dramatic reallocation of AI workloads but a steady, contract-by-contract tightening of data-residency and retention clauses — the kind of shift that shows up first in procurement documents and only later in where capacity gets built.
The honest caveat is that fines of this size have repeatedly failed to change platform behaviour on their own, because the revenue attached to the underlying data practice can exceed the penalty. The more consequential outcome is usually the remediation order that accompanies the money — what a company must stop doing, and by when. Until Google details what changes, the $463 million is a price, not yet a redesign.
Background
The European Union operates the world’s most consequential data-protection regime, built on the principle that personal data may only be collected for a specified purpose, with a lawful basis, and kept no longer than necessary. Enforcement is carried out by national authorities and courts, and penalties are calculated against a company’s global turnover rather than its local revenue — which is why actions against the largest technology firms produce figures in the hundreds of millions or billions. Location data has been a recurring focus because it is collected continuously, is rarely the service a user actually asked for, and is unusually revealing when aggregated.
Alphabet’s Google has been the subject of repeated European regulatory action across competition, advertising and data protection over the past decade. It is simultaneously one of the largest operators of data-centre and cloud infrastructure in Europe, and one of the most aggressive builders of artificial-intelligence capacity — which places the same corporate entity on both sides of this story: the regulated handler of consumer data, and the supplier of the infrastructure that enterprises use to process their own.
Building a large data center is closer to building a small power station than an office block, and running late is expensive.
Green Mountain, a company that builds data centres in Norway, Germany and the United Kingdom using only renewable electricity, has brought in a software firm called Buildots. Cameras worn around the site and drones flying over it record what has actually been built, and the software compares that against the plan and flags delays early.
It is already running on two sites, in Germany and Norway, that together will draw about as much electricity as a small city.
Buildots announced on September 22 an enterprise agreement with Green Mountain, the Norwegian developer and operator of sustainable data centers, covering the rollout of its construction intelligence platform across Green Mountain’s project portfolio. The platform is already live on two projects totalling more than 100 megawatts, in Frankfurt, Germany and Stavanger, Norway, with further expansion planned.
Under the agreement, Buildots covers much of the project lifecycle, from superstructure through to fitout, combining 360-degree site cameras with drone imagery to generate progress data and early risk insights. Neither company disclosed contract value, duration, or the number of future projects covered.
Executive Summary
The commercial substance here is small in dollar terms — no value was disclosed — but the operational signal is specific. A developer that markets itself on energy efficiency and 100% renewable power has decided that the next efficiency to chase is delivery: getting megawatts finished on the date it promised them.
That reflects where the constraint has moved. In a market where land, grid connections and equipment lead times are all contested, the one variable a developer directly controls is how well it executes a build it has already financed and permitted. Green Mountain’s Chief Design & Construction Officer, David Guth, framed the outcome as a repeatable “playbook” — consistent methodology, consistent visibility, consistent outcomes — rather than a one-project tool.
For the wider sector, the deal is a data point on the adoption curve of construction intelligence: Buildots CEO Roy Danon described the category as moving “from something experimental to a must-have” among developers and colocation providers. That is a vendor’s characterisation of its own market, and the release offers no adoption statistics to test it — but the named customer list behind it is not trivial.
Why the Schedule Became the Scarce Input
A data center megawatt earns nothing until it is energized and handed over. That sounds obvious, but it has become the binding economic fact of the current buildout. Chips, transformers, switchgear and grid capacity all have queues, and a developer can do relatively little about any of them once orders are placed. What a developer does control is the twelve to thirty months between breaking ground and commissioning — and that window is where schedule slips accumulate quietly, discovered at milestone reviews rather than in the week they happen.
The mechanism Buildots sells is the compression of that discovery lag. Helmet-mounted 360-degree cameras and drone flights capture what physically exists on site; the platform compares that record against the plan and the schedule and surfaces deviations as progress data rather than as a contractor’s monthly narrative. Applied from superstructure through fitout, as this agreement specifies, it covers the phases where data center schedules actually go wrong: not the shell, but the long tail of electrical rooms, mechanical plant and cable containment that determines the energization date.
At the 100-plus megawatts already live in Frankfurt and Stavanger, the arithmetic of that lag is what makes the deal interesting. A month of undetected slip on a 100 megawatt asset is a month in which financed, permitted, powered capacity produces nothing — and, in a tight market, a month a tenant waits. The release says the platform gives Green Mountain “greater control over its capital investment” and “early visibility to protect the delivery dates that matter most.” Stripped of marketing register, that is a claim about converting schedule risk from a narrative into a measurement. The parties affected extend past the owner: general contractors whose progress claims are now checked against imagery, subcontractors sequenced around a shared view of site reality, and lenders and tenants who price a delivery date they cannot themselves inspect.
Sustainability Discipline Moves Upstream Into the Build
Green Mountain’s public identity rests on operations: 100% renewable power, world-leading energy efficiency, sites in Norway where cool climate and hydropower do much of the work. The release makes an explicit move to extend that discipline backwards into construction, with Danon tying the platform to “sustainability goals through more efficient construction.”
The logic is sound in principle. Construction waste, rework and idle plant carry a real carbon and cost load, and a facility’s embodied emissions are fixed before a single server racks up. But the release supplies no measurement of it — no tonnes avoided, no rework percentage, no reduction in schedule variance. Operational efficiency at Green Mountain is a metered, auditable claim; construction efficiency here is, so far, an argument rather than a number.
The geography is worth noting on its own. Running the same methodology across Stavanger and Frankfurt means applying one toolset to two very different delivery environments — Nordic sites with abundant renewable power and lower land pressure, and a Frankfurt market that is among Europe’s most constrained on power and permitting. A playbook that survives both is more portable than one proven in a single jurisdiction.
What This Proves About Construction Intelligence — and What It Doesn’t
Buildots describes itself as “emerging as a gold standard” for schedule-critical builds and cites Turner Construction, JE Dunn, Intel, HOCHTIEF and Bouygues among its users. Those are substantial names, and the company separately announced a $130 million funding round explicitly tied to the data center buildout. That establishes credible commercial traction.
It does not, by itself, establish the category shift Danon claims. “From experimental to must-have” is a statement about an entire buyer population, and no penetration rate, retention figure or measured outcome accompanies it. Enterprise agreements of this kind are also frequently structured as portfolio frameworks with per-project call-off — meaning “across the portfolio” can describe an option to deploy rather than a committed deployment. The two live projects are the substantiated part; “further expansion planned” is an intention.
For buyers evaluating this class of tool, the honest read is that the evidence base is one of adoption, not yet of published outcome. The question to put to any vendor in the space — Buildots included — is not how many sites are instrumented but how much schedule variance narrowed after instrumentation, measured against a comparable baseline. Neither company has published that figure for these projects.
Background
Green Mountain designs, builds and operates data centers in Norway, Germany and the United Kingdom, marketing itself on 100% renewable power and world-leading energy efficiency. Its Norwegian footprint takes advantage of hydropower and a cool climate, both of which lower the energy cost of cooling — historically one of the largest operating expenses in the sector. Its expansion into Frankfurt places it in one of Europe’s most power- and permit-constrained data center markets.
Buildots sells into a construction industry that has long struggled to measure in-progress work objectively, relying on contractor-reported percentages that are difficult for owners to verify. Its platform captures site conditions through 360-degree cameras and drones and compares them to design and schedule data. The company has raised $130 million and counts Turner Construction, JE Dunn, Intel, HOCHTIEF and Bouygues among its users, and has been expanding toward large, schedule-critical projects — of which AI-era data centers are now among the most numerous.
Only about one in three people in Africa used the internet last year. In the rest of the world, it is closer to three in four.
The United States government’s overseas investment arm plans to buy a stake in WIOCC Group, a company that lays internet cables under the sea and across land in more than 30 African countries, and runs the buildings that store and move the data. A pan-African development bank and a Saudi infrastructure investor are backing it too.
The price has not been made public, and lawmakers in Washington must be notified before the deal can close.
WIOCC Group announced on September 22, 2026 that the U.S. International Development Finance Corporation (DFC) intends to join Africa Finance Corporation (AFC) and Vision International Investment Company (Vision Invest) as an investor in the company. The announcement, issued from New York on behalf of AFC, followed a roundtable AFC hosted on the sidelines of the Global Africa Business Initiative’s Unstoppable Africa event, held alongside the UN General Assembly.
The combined investment is earmarked for WIOCC Group’s terrestrial and subsea networks, its data centre capacity and its broader digital infrastructure platform. WIOCC Group operates in more than 30 African countries and has been trading since 2008. No investment amount, ownership percentage or valuation was disclosed, and DFC’s participation remains subject to further steps before commitment and closing, including congressional notification.
Executive Summary
The transaction puts three very different pools of capital on one cap table: a pan-African multilateral investor, a Saudi infrastructure developer, and the investment arm of the United States government. What they are buying into is not compute. It is the layer underneath it — undersea cables, long-haul and metro fibre, and carrier-neutral data centres, meaning facilities that host any operator’s equipment without competing for their customers.
The framing in the announcement is explicitly about artificial intelligence and cloud demand, but the numbers cited describe a access problem rather than a processing one. The International Telecommunication Union figure quoted puts 35.7% of Africa’s population online in 2025 against a 73.6% global average — a gap of nearly 38 percentage points. UNCTAD’s projection of a US$4.8 trillion global AI market by 2033 is paired, in the same announcement, with that body’s warning that AI capability and digital infrastructure remain concentrated in a small number of countries and companies.
What matters for the industry is the reclassification. Fibre, subsea systems and data centres are being underwritten here the way roads, ports and power plants are — as strategic infrastructure with a long payback, rather than as telecom credit. What the announcement does not do is put a figure, a capacity number, a customer or a date against that intent.
Why the Constraint Is Reach, Not Chips
The headline number in this announcement is 35.7%. That is the share of Africa’s population the ITU recorded as using the internet in 2025, against a global average of 73.6%. Hold that next to what the money is actually going toward: terrestrial networks, subsea cables, data centre capacity, and the platform that ties them together. Accelerators, GPUs and compute clusters appear nowhere in the list. That is not an oversight — it is the sequence in which this infrastructure has to be built.
The mechanism is straightforward once you separate the layers. AI capacity is only useful if traffic can reach it cheaply and quickly. A cloud region needs a subsea landing station to reach the rest of the world, long-haul fibre to carry traffic inland, metro fibre to reach the cities where users are, and neutral interconnection points where networks meet and exchange traffic. Where those layers are thin, compute placed on the continent has an expensive route to its users, and users have an expensive route to compute hosted elsewhere. Latency and transit cost, not silicon availability, are what determine whether an African workload runs locally or in Europe.
This is not a claim that chips are irrelevant. It is a claim about where capital binds first, and about who is affected downstream. Carriers buying wholesale capacity, enterprises pricing cloud egress, content and streaming providers weighing whether to cache locally, and hyperscalers deciding whether a market justifies a region all sit behind the same constraint. WIOCC’s own chief executive frames the use of proceeds as accelerating data centre deployment and consolidation, expanding open-access terrestrial fibre, and investing in new subsea assets — three transport-and-hosting moves, no compute commitments among them.
Development Finance Stops Pricing Fibre as Telecom Risk
AFC chief executive Samaila Zubairu made the reclassification explicit: just as transport corridors enable trade and energy networks power industry, fibre, data centres and subsea cables are now essential infrastructure. That is the language of a project finance desk, not a telecom credit committee, and it matters because the two price risk very differently. AFC has invested over US$19 billion across 36 African countries since 2007 in energy, natural resources, heavy industry, transport and telecommunications; this deal moves digital infrastructure firmly into the first category rather than leaving it in the last.
The economics behind that shift are unforgiving and well understood by anyone who has financed a cable. Subsea systems and data centres carry almost all their cost up front and earn it back over fifteen to twenty-five years, and an open-access wholesale model — selling capacity to all comers rather than bundling it into retail service — earns slowly by design, because its margin discipline is the point. Equity from investors who can tolerate that horizon is typically what makes commercial debt behind it financeable at all. An expanded shareholder base of development, sovereign-linked and multilateral capital is therefore a balance sheet event before it is a construction event.
The practical consequence for the market is competitive rather than charitable. Open-access capacity lowers the cost for every other operator to reach the same markets, which compresses wholesale transit pricing and improves the business case for anyone contemplating a landing or a region. That is a benefit to WIOCC’s competitors as much as to WIOCC, and it is the standard argument for why public capital is placed at this layer.
The Hyperscaler Sentence, and What It Does Not Commit To
The most quoted line in the announcement will be DFC’s. Conor Coleman, the agency’s chief of staff and head of investments, said the platform will support American and allied partner companies seeking to grow in the market, including U.S. hyperscalers and the broader American technology ecosystem. DFC describes itself as central to U.S. economic statecraft, with a mandate that includes securing supply chains. Read plainly, that is industrial policy applied to connectivity, and the three-flag investor group — African, Saudi and American — is itself the story the parties want told.
It is worth being precise about what that sentence establishes and what it does not. Substantiated by the announcement: the investor set, the intent, the categories of spend, WIOCC’s presence in more than 30 countries, and the market context the ITU and UNCTAD figures supply. Not substantiated: any hyperscaler customer, contract, anchor tenancy or committed capacity. A stated willingness to serve a class of customer is a demand thesis, not demand.
Nor is this closed. DFC’s participation is explicitly subject to further steps before commitment and closing, including congressional notification. Announced intent from a government investor carries real signalling weight for co-investors and lenders, but it is not funded capital, and the distinction is the one buyers and counterparties should hold onto until a closing is confirmed.
Background
WIOCC has operated in Africa since 2008, building out from wholesale subsea and terrestrial capacity into a group that now spans metro fibre, carrier-neutral data centres and open-access platform services across more than 30 countries. The open-access model it runs — selling capacity to all operators on equal terms rather than reserving it for a retail arm — is the structure most African markets have leaned on to bring down the cost of international bandwidth.
Africa Finance Corporation was established in 2007 as a multilateral infrastructure investor and now counts 48 member countries, with more than US$19 billion invested across 36 African countries in energy, natural resources, heavy industry, transport and telecommunications. Vision Invest is a Saudi infrastructure investment and development company active in energy transition, digital infrastructure, transport and logistics, and social infrastructure, with holdings across five continents. The addition of DFC brings the investment arm of the U.S. government alongside both.
Computers that train artificial intelligence have to talk to each other constantly, and today most of that traffic still travels over copper wiring that burns electricity and fades over distance.
Marvell has shown chips that move the traffic as light instead, built on the most advanced chipmaking process now available. One demonstration moves data fast enough to copy well over a thousand high-definition movies every second.
The point is not speed. Light uses less power per unit of data, and electricity is the hard limit on how big these computing sites can get.
Marvell Technology has demonstrated optical interconnect silicon built on a 2nm process for AI data centers, including a co-packaged optics engine running at 102.4 terabits per second and supporting 200G and 400G per electrical lane. 24/7 Wall St. reported on September 21 that the company described the work as an industry first and said it is “pushing the envelope of what is possible in optical tech,” with shares trading around $248 in Monday’s first hour, up 1.5% on the session.
The demonstration sits on top of a quarter Marvell has already reported: record fiscal second-quarter revenue of $2.739 billion, up 37% year over year, with data center revenue of $2.17 billion, up 46%, now 79% of the total. Guidance filed August 27 put the current quarter at $3.150 billion plus or minus 5%, and raised the fiscal 2027 and fiscal 2028 outlooks. Management has scheduled an Investor Day for October 6 in New York.
Executive Summary
The headline number is 102.4 Tbps of co-packaged optics, but the operational story is the denominator. Co-packaged optics — CPO — moves the optical engine off the front panel of a switch and onto the same substrate as the switch or accelerator silicon. That shortens the electrical path a signal has to travel before it becomes light, and shortening that path is primarily a way to spend fewer watts per bit, not a way to go faster.
That distinction matters because a rack has a fixed power envelope set by the building, not by the silicon roadmap. Every watt the network consumes inside that envelope is a watt unavailable to the accelerators the network exists to connect. As Marvell steps its lane rates from the 800G generation deployed today, past the 1.6T generation now ramping, toward 3.2T, copper’s usable reach shrinks and more of the interconnect budget migrates to optics — which is exactly where the energy cost concentrates.
The demo also gives silicon substance to forecasts management has already made. On the earnings call, CEO Matt Murphy said “the magnitude of our scale-up optics opportunity next year is much larger than we thought just a quarter ago,” and that Marvell expects scale-across bandwidth requirements to exceed 10 times current front-end data center interconnect networks. What a demonstration does not establish is yield, production timing, field serviceability or a named customer.
Optics Stopped Being a Bandwidth Problem and Became a Power Budget
Here is the mechanism the 102.4 Tbps figure is really about. In a conventional switch, optical modules plug into the faceplate, and the switch chip’s electrical lanes have to drive signals across the package, the circuit board and a connector before they reach the laser. Each of those hops costs energy and frequently requires retiming circuitry to clean the signal up. Co-packaged optics collapses that journey by placing the optical engine on the same substrate as the switch or accelerator die — in Marvell’s framing, shortening data pathways and cutting energy per bit.
The lane rates are what force the issue. By demonstrating 200G and 400G per lane, Marvell is aiming past the 800G optics deployed across hyperscale fleets today and the 1.6T generation now ramping, at the 3.2T generation beyond it. As per-lane signalling rates climb, the distance a passive copper cable can carry a clean signal falls, which is why the release notes copper runs out of reach at the speeds these clusters require. Once copper is out, the interconnect is optical by default — and its power draw becomes a line item in the rack budget rather than a rounding error.
Who this lands on is specific. A facility commits a fixed number of megawatts to a hall; a rack design commits a fixed number of kilowatts to a cabinet. If the network layer connecting tens of thousands of accelerators across multiple racks consumes a smaller share of that envelope, the operator either fits more compute per rack or fewer racks per hall. That arithmetic is the reason a demo with no customer attached still moves the roadmap conversation for switch vendors, optical module suppliers, rack integrators and the people underwriting the power contracts underneath them.
A 79% Data Center Mix Makes the Optical Roadmap the Whole Company
Marvell’s reported quarter explains why an engineering milestone carries this much weight. Revenue of $2.739 billion, up 37% year over year, came with data center revenue of $2.17 billion, up 46% and equal to 79% of the total. Non-GAAP earnings were $0.94 per share, and the company guided the following quarter to $3.150 billion plus or minus 5% with earnings of $1.10 plus or minus $0.05, while raising its fiscal 2027 and fiscal 2028 outlooks in the August 27 filing.
A business that derives roughly four dollars in five from data centers has effectively converted itself into a single-cycle instrument. That is an advantage while hyperscale capital expenditure expands and a concentrated exposure when it does not. It also means the optical and custom silicon roadmaps are not adjacent product lines — they are the earnings model, which is why a 2nm demonstration is treated as corroborating evidence for guidance rather than as a lab result.
The market has priced that trajectory rather than waiting on it. The stock has advanced 192% year to date, carries a market capitalisation near $217.4 billion and a forward price-to-earnings multiple around 58, against an average analyst target of $289 with 39 buy or strong-buy ratings and five holds. A multiple set that far ahead of current earnings is a statement about execution risk: the roadmap has to arrive roughly on the schedule management has sketched, and the October 6 Investor Day is where that schedule gets specified.
What Being First Actually Buys — and What It Costs
The optical push runs alongside a custom silicon business that has been given an unusual structure. Management disclosed an expanded partnership with Google that includes a warrant allowing Google to acquire up to 7% of Marvell’s shares, tied to revenue milestones, and Murphy said custom revenue should more than double in fiscal 2028. Linking a customer’s equity upside to the volume it buys aligns incentives efficiently, but it also converts a commercial relationship into potential dilution for existing holders, on terms whose triggers Marvell has not published. A separate expanded NVLink Fusion arrangement with NVIDIA positions Marvell in scale-up switching, and the earlier Celestial AI photonic fabric acquisition, closed February 2, 2026, and the XConn Technologies chiplet connectivity deal, closed February 10, 2026, supply the building blocks.
Against Broadcom, the incumbent in high-speed networking silicon, first-to-2nm is a claim on the next design cycle rather than on current sockets. Hyperscalers qualify interconnect years ahead of deployment, so the practical value of a demonstration is optionality in roadmap reviews happening now for clusters that power up later. It converts into revenue only if the process yields at volume, the thermal and reliability behaviour survives qualification, and a named platform adopts it.
The demand side of that equation is funding itself in size, if not always for named projects: Amazon’s Form 8-K filed September 14 records the close of £4.242 billion of notes maturing between 2029 and 2045, with net proceeds of roughly £4.235 billion and no specific use of proceeds identified in the filing. Capital of that scale is available to the companies buying interconnect; whether it reaches co-packaged optics specifically is the question Marvell has not yet answered with a customer name.
Background
Marvell Technology designs custom silicon and networking chips for hyperscale data centers. Its portfolio spans custom AI accelerators built to a single customer’s specification, electro-optics, the digital signal processors that drive high-speed optical links, and data center interconnect products that carry traffic between facilities. Over the past two years the company has reweighted itself decisively toward that market, with the data center segment now accounting for 79% of revenue.
Interconnect has become one of the more consequential bottlenecks in AI infrastructure. Training and inference clusters link tens of thousands of accelerators, and the network that connects them has to scale in step with the compute — first inside a rack, then across racks, then between buildings. Marvell has assembled capability for that through acquisition as well as internal development, closing the Celestial AI photonic fabric deal on February 2, 2026 and the XConn Technologies chiplet connectivity deal on February 10, 2026, while expanding partnerships with NVIDIA on scale-up switching and with Google on custom silicon. Its principal competitor in high-speed networking silicon is Broadcom.
A company best known for mining Bitcoin has signed a deal worth 350 million US dollars over five years to rent out computing power for artificial intelligence.
The company, HIVE Digital Technologies, runs warehouses full of computers, largely on renewable electricity. It is steadily pointing those machines at artificial intelligence work instead of Bitcoin.
Why it matters: the deal promises roughly 70 million dollars a year, about double what HIVE’s artificial intelligence business brings in today. The hard part now is getting the machines installed and running on schedule.
scanx.trade reported that BUZZ High Performance Computing Inc. (“BUZZ HPC”), the wholly owned subsidiary of HIVE Digital Technologies Ltd. (TSX: HIVE; Nasdaq: HIVE), has signed a $350 million, five-year AI cloud services agreement — a contract that would supply GPU compute capacity, the specialised processors used to train and run artificial-intelligence models, on a multi-year committed basis rather than by the hour.
The deal arrives in an unusually dense month of disclosure for HIVE. In an 8-K filed September 16, the company told investors that on August 21, 2026 it exceeded $1 million in combined average daily revenue from Bitcoin hashrate and GPU cloud services, with the GPU cloud business contributing roughly $100,000 of that per day. A second 8-K, filed September 17, disclosed a partnership between BUZZ HPC and Vancouver-based applied-AI firm ProCogia announced on September 16.
Executive Summary
A $350 million contract with a five-year term is a different financial object from a GPU rental business. Spot and short-term GPU capacity is priced like a commodity and repriced constantly; a dated multi-year agreement converts that capacity into something closer to a lease — a revenue line with a start date, an end date and a counterparty obligation behind it. For HIVE, which spent years as a Bitcoin miner monetising renewable power through a volatile, unhedged commodity, that distinction is the whole point of the BUZZ HPC strategy.
The scale is worth measuring carefully in both directions. Against HIVE’s disclosed GPU cloud run-rate of about $100,000 per day — roughly $36.5 million annualised, per the company’s September 16 filing — the contract’s straight-line average of $70 million a year is roughly double the entire AI segment as it stands. Against the whole company, which HIVE says has been running above $1 million in combined daily revenue, the same contract is smaller than a single year of current total revenue. It is transformative for one engine and incremental for the business overall.
What changes with a signed multi-year contract is the location of the risk. Demand risk falls; delivery risk rises. HIVE’s own forward-looking language in the September 17 filing names the exposure precisely: the ability to convert “signed total contract value into recognized revenue on the timing anticipated,” alongside “delays in equipment delivery or commissioning” and “capital availability.” Those are procurement, construction and financing problems, not sales problems.
The Arithmetic: What $350 Million Means in a $100,000-a-Day Business
Start with the two numbers side by side. HIVE disclosed in an 8-K filed September 16 that since August 21, 2026 its GPU cloud operations generated approximately $100,000 in average daily revenue. Annualised on a straight line, that is roughly $36.5 million. A $350 million agreement spread evenly across five years implies about $70 million a year. The contract, on its face, is worth close to twice the current annual run-rate of the business it is meant to feed.
That ratio is the operationally interesting part, because a contract that large relative to an existing book cannot be served out of spare capacity. It has to be built. Real contracts also rarely run straight-line: multi-year compute agreements typically ramp as racks energise, which means the revenue curve is back-weighted and the capital spend is front-weighted. Between signature and recognition sit accelerator delivery slots, power energisation, cooling and network commissioning, and the working capital to pay for hardware before the customer pays for capacity. HIVE says as much in its own risk disclosure, listing equipment delivery delays, commissioning delays and capital availability as the factors that could break the timing.
Who is affected downstream is concrete. Accelerator suppliers and system integrators gain a scheduled order rather than a speculative one. Electrical contractors, cooling vendors and interconnection counterparties — the utilities and operators who connect a site to the grid — get a date to work against. And lenders get what they have been asking for from this asset class: a named term and an amount, which is what makes GPU fleets financeable at something other than venture pricing. The flip side is that a dated contract removes the operator’s option to slip. A miner can idle machines when economics turn; a contracted AI host that misses a commissioning window is in breach, not merely idle.
Dual Engines, Different Risk: 12 Bitcoin a Day Versus Contracted Compute
HIVE’s September 10 release, furnished to the SEC on September 16, laid out the current mix plainly: approximately 12 Bitcoin per day, which the company put at roughly 2% of global Bitcoin network production, plus about $100,000 a day from GPU cloud, together exceeding $1 million in combined average daily revenue. On those figures the AI business is still well under a tenth of the top line. The $350 million contract does not reverse that overnight — but it changes the character of the revenue it adds.
Mining revenue is spot revenue. It moves with Bitcoin’s price, with network difficulty, and with the scheduled supply reductions built into the protocol, and no amount of operational skill hedges those. Contracted compute revenue is dated and, depending on structure, may be substantially insensitive to what the spot GPU market does over the same five years. That is the trade the entire miner-to-AI cohort is attempting: exchange high-velocity, high-variance cash generation for lower-variance, longer-duration cash flows that a lender will underwrite. HIVE’s own framing in the September 10 release — Bitcoin as a “high-velocity cash-generating foundation” funding a “recurring, contracted AI infrastructure business” — is an accurate description of the mechanism, and it is also an admission that the second engine is not yet self-financing.
The portfolio reallocation is already visible. HIVE said its Swedish Bitcoin mining operations represented less than 5% of global daily revenue during August and that it intends to wind down mining there while evaluating repurposing that infrastructure for HPC and AI. The candour of that number matters: it tells you the wind-down is a low-cost decision, not a painful one. It also warrants a caution the release does not dwell on. Repurposing a mining hall for AI is not a software change. Mining sites are built for cheap power, high temperature tolerance and thin redundancy; accelerated compute for enterprise customers needs denser power per rack, liquid or high-capacity cooling, carrier-grade networking and the uptime engineering implied by the Tier-III designation HIVE cites for BUZZ HPC. Some megawatts convert cheaply. Others are effectively a new build on an old plot.
Sovereign Compute Is the Sales Motion, Not Just the Slogan
The ProCogia partnership disclosed on September 17 shows how BUZZ HPC intends to fill capacity beyond a single anchor contract. ProCogia will procure dedicated GPU capacity in BUZZ HPC’s Canadian data centres to run its ZeroBoxx AI framework and vertical products — CallYeah, a voice-AI tool for healthcare providers, and PolyKode, a code-migration model for regulated environments — and will in turn act as a preferred applied-AI services partner to BUZZ HPC’s own customers, with the two intending to co-sell into the United States and Europe.
Read commercially, that is a channel strategy aimed at a specific buyer problem. “Sovereign” compute means infrastructure where the data and the processing stay inside a defined legal jurisdiction, which matters to healthcare, financial services and public-sector buyers governed by data-residency rules. Those buyers are also the ones least able to assemble an AI stack themselves, so bundling infrastructure with an applied-AI delivery partner addresses the real obstacle — nobody to build the thing — rather than just the rack. The hire of Mark Volk as BUZZ HPC’s Senior Vice President, Revenue, announced in the same September 10 release, fits the same thesis: an enterprise and government sales motion is a different discipline from selling hashrate.
The honest caveat is that the partnership discloses no dollar value, no committed capacity and no term, which places it in a different category from the $350 million agreement. Executive Chairman Frank Holmes referenced “a 400 MW coast-to-coast Canadian pipeline” — 400 megawatts being, very roughly, the electricity draw of a mid-sized city — but a pipeline is not energised capacity, and the release does not break it into secured power, sites under construction and sites under evaluation. Competition is real, too: the major cloud providers have been standing up in-country regions with sovereignty commitments of their own. BUZZ HPC’s argument is price and availability for Canadian companies that cannot get either, which is a defensible niche — and one that a five-year contract, if it converts on schedule, would do more to prove than any amount of positioning.
Background
HIVE Digital Technologies Ltd. (TSX: HIVE; Nasdaq: HIVE; BVC: HIVECO), headquartered in San Antonio, Texas, began as a Bitcoin miner built around low-cost renewable power and data centre operations. Its stated model is to convert electricity into compute and then sell that compute — historically as hashrate for the Bitcoin network, and increasingly as GPU capacity for artificial-intelligence workloads through its wholly owned subsidiary BUZZ High Performance Computing Inc.
That pivot is an industry-wide pattern. Miners already control three scarce inputs for AI infrastructure — secured power, energised sites and operational experience running large fleets of hot, dense hardware — while facing a commodity revenue stream that is volatile and periodically halved by protocol design. Converting megawatts to AI hosting trades that volatility for contracted, longer-duration cash flows, but requires heavier capital per megawatt and a customer base that demands uptime guarantees, data-residency assurances and enterprise service levels that mining never did.
The computer chips that run artificial intelligence get so hot that blowing cold air over them is no longer enough. The fix is to pipe liquid right up to the chip.
OData, a data center operator active across Latin America, is spending 630 million dollars to add that liquid plumbing to two buildings it already runs, one in Brazil and one in Mexico. The work has started.
Why it matters: the region can host the newest, hottest artificial intelligence machines by upgrading buildings that already exist, instead of waiting years for new ones to be built.
Data Center Dynamics reported that OData is installing DeltaFlow, a proprietary liquid cooling system developed by its parent company Aligned Data Centers with the power and cooling vendor Munters, at its SP04 data center in Brazil and its QR03 facility in Mexico. The project is priced at $630 million across the two sites, and phase one installation and testing is already underway at both.
Both facilities already run Aligned’s Delta Cubed air cooling system. DeltaFlow will let them operate as hybrid halls — air cooling for conventional servers, liquid cooling for the high-density racks that AI accelerators require. The DeltaFlow units are manufactured at Munters’ plant in Araucária, in the Brazilian state of Paraná.
Executive Summary
This is a retrofit, not a groundbreaking. OData is adding direct liquid cooling — coolant carried to a plate sitting on the processor itself, rather than chilled air pushed across a room — to two data centers that are already built, already powered and already in service. That distinction is the whole economic argument. Liquid cooling is what allows a rack to draw the tens of kilowatts that AI accelerators consume, and the fastest route to that capability is almost always a building that already exists.
The $630 million figure is unusually large for a project described in terms of cooling, which suggests scope well beyond swapping chillers: retrofits of this type typically pull in electrical distribution, floor loading, piping, heat rejection and controls. The release does not break the number down, and it does not state how much liquid-cooled capacity the spend produces.
The second thread is manufacturing. DeltaFlow is built at a Munters factory in Paraná, inside Brazil. Long-lead mechanical equipment has been one of the more stubborn schedule constraints in data center construction, and supplying a Brazilian deployment from a Brazilian plant removes ocean freight and customs from that path.
The Retrofit Is the Point
OData’s two target sites are not new. The release states that both already operate Aligned’s Delta Cubed air cooling, and that DeltaFlow will enable them to run as hybrid air-and-liquid facilities catering to high-density AI racks. In plain terms: the buildings were designed for servers that a stream of cold air could keep in range, and they are being upgraded for servers that it cannot. A modern AI accelerator rack concentrates heat in a volume where moving enough air becomes impractical and expensive long before it becomes impossible, which is why coolant is brought to the chip instead.
What that changes operationally is the pacing item. A new data center is gated by land, permits, construction and — most painfully in the current cycle — a grid connection, the queue to get utility power delivered to the site. A retrofit inherits all of those. The gate becomes mechanical installation and commissioning inside a shell that already has power, floor, network and security in place. That is measured in a fit-out schedule rather than a development schedule, and it is the reason a hybrid upgrade can put dense capacity into a market faster than any greenfield project competing for the same customer.
The people this reaches are buyers in Brazil and Mexico who want to train or serve AI models on machines physically located in the region — for latency, for data residency, or simply because capacity elsewhere is spoken for. The release confirms the capability is coming and that work has begun. It does not say how many megawatts of liquid-cooled capacity results, or what rack density in kilowatts the halls will support, so the size of that advantage cannot yet be measured from the outside.
$630 Million Is a Building Number, Not a Chiller Number
The figure is given for the project across both sites, without itemisation. It is worth being precise about what a cooling retrofit actually involves, because the headline number only makes sense at that scope. Introducing liquid cooling into a live air-cooled hall generally touches the electrical distribution feeding denser racks, the structural loading of the floor, the piping and manifolds that carry coolant, the heat rejection plant outside the building, the water or refrigerant management behind it, and the control systems that have to run two cooling regimes at once without either interfering with the other — all while the existing customers stay online.
What is substantiated here is a capital commitment of a stated size, a named technology, two named facilities and work that has begun. What is not substantiated is the unit economics. Without a capacity figure, there is no cost per megawatt, and cost per megawatt is precisely the number that would tell a buyer or an investor whether upgrading beats building. That is a disclosure question for OData and Aligned, not a flaw in the underlying project.
There is a strategic read in how the system is sourced. Aligned designed DeltaFlow itself, with Munters, rather than buying merchant coolant distribution units off a catalogue. For a group operating across several countries, a single in-house platform means common spares, common training and a repeatable commissioning process — real advantages when you are staffing facilities in four national markets. The trade-off is concentration: the platform’s roadmap, and the factory behind it, become a dependency.
The Factory in Paraná Is Part of the Infrastructure Story
DeltaFlow is manufactured at Munters’ plant in Araucária, Paraná, in southern Brazil. That detail does more work than it first appears to. Across the industry, the binding constraint on data center schedules has repeatedly been long-lead mechanical and electrical equipment rather than chips or capital. Supplying a Brazilian deployment from a Brazilian factory takes ocean freight, port handling and import clearance out of the critical path for at least one of the two sites.
Mexico is still a cross-border shipment, so the benefit is not symmetrical. But the same production line and the same platform serve both, and Munters’ Keith Dunnavant framed the company’s global manufacturing footprint as what allows it to support Aligned and OData as they expand across Latin America. A manufacturing presence in-region also tends to bring service technicians and spare parts inventory with it, which matters more for liquid systems than air ones: a leak or a pump failure in a liquid loop is a faster-moving problem than a failed fan.
Neither company has quantified the plant’s output or said whether any of it is reserved for this project, so the supply-chain advantage is directional rather than measured.
What It Signals About Latin American Colocation
OData was founded in 2015 by the Brazilian private equity firm Patria Investments and acquired by Aligned Data Centers in May 2023. It runs facilities in Chile, Colombia, Brazil and Mexico with more in development. This deployment is, in effect, technology transfer from a North American operator into a regional platform it now owns — the same pattern the company followed with Delta Cubed on the air side.
OData chief executive Ricardo Alário described the goal as establishing DeltaFlow as the market benchmark for liquid cooling in the region. That is an ambition rather than a disclosed market position, and it should be read as one. What the announcement does establish is concrete and checkable: a stated sum, two identified sites, a named manufacturing source and construction in progress. What it does not establish is demand. No anchor customer, pre-lease or contracted capacity has been disclosed, and a retrofit can be built ahead of committed tenants as easily as behind them.
For competing operators in the same two markets, the practical effect is that rack density moves onto the specification sheet. Once one regional provider can quote liquid-cooled halls, requests for proposals start asking every provider what they can support per rack — and the answer for an air-only site is a hard ceiling that no amount of floor space compensates for.
Background
OData was founded in 2015 by Patria Investments, a Brazilian private equity firm, and grew into one of the region’s larger colocation platforms — operators that lease space, power and cooling to companies that prefer not to run their own facilities. Aligned Data Centers, a US operator known for developing its own cooling technology rather than buying it, acquired OData in May 2023. OData now runs facilities in Chile, Colombia, Brazil and Mexico with further sites in development, and has been the route by which Aligned’s in-house systems reach Latin American markets: the same two data centers receiving DeltaFlow already run Aligned’s Delta Cubed air cooling.
The wider context is the shift in what a data center has to physically accommodate. For two decades, racks drew a few kilowatts each and chilled air sufficed. AI accelerators have pushed rack power an order of magnitude higher, past the point where air can economically carry the heat away. That has made cooling — historically an engineering afterthought to power and space — a gating factor on whether a given building can host AI workloads at all, and it has turned retrofits of existing, already-powered facilities into one of the faster routes to new AI-capable capacity.
A small energy and data center developer, New Era Energy & Digital, agreed to buy electricity from Vistra, one of the largest private power generators in the United States, for the next twenty years.
The amount is 207 megawatts — roughly the electricity a small city uses, running day and night.
New Era’s share price jumped 16 percent on the news.
Why it matters: computer warehouses that run artificial intelligence need enormous, uninterrupted power. Locking in a supply for two decades is increasingly what banks require before they will lend the money to build one.
New Era Energy & Digital has signed a 20-year power purchase agreement with Vistra Corp. covering 207 megawatts of electricity, the company announced in a release distributed through GlobeNewswire. A power purchase agreement, or PPA, is a long-term contract under which a buyer commits to purchase a set volume of electricity from a generator, usually at a pre-agreed price structure.
Investing.com reported that New Era’s stock rose roughly 16% on the announcement. Vistra, an NYSE-listed independent power producer with a large fleet of generating plants and a retail electricity business, sits on the supply side of the contract. The announcement disclosed the counterparty, the contract term and the contracted capacity; it did not attach a price, a delivery start date or a named end user for the power.
Executive Summary
Two numbers define this deal: 207 and 20. The first is the capacity — enough to power a campus-scale data center, not a pilot. The second is the term, and it is the one that matters more to anyone financing digital infrastructure. Twenty years sits at the long end of the range for corporate power contracts, which more commonly run ten to fifteen years.
The significance is structural. Over the past two years, the binding constraint on new AI-capable data center capacity has moved from chips to land to interconnection to, now, contracted electricity. A developer that can show a lender a two-decade supply agreement with a creditworthy generator has converted its single largest and most volatile operating input into something closer to a fixed cost. That is the document that makes a construction loan underwritable.
For Vistra, the transaction is an example of the trade that has re-rated the independent power sector: selling long-dated output to digital load instead of leaving it exposed to wholesale market swings. For New Era — a far smaller company, judging by the size of the share-price reaction — it is a credential as much as a commodity purchase.
Why the Term Sheet, Not the Megawatts, Clears the Financing
Lenders to data center projects underwrite two things above all: revenue certainty on the customer side and cost certainty on the input side. Power is typically the largest recurring operating expense in a high-density computing facility, and in wholesale markets it is also the most volatile. A project that buys electricity at spot or on a short rolling contract carries an open-ended cost exposure across a 15-to-20-year debt tenor — precisely the mismatch that credit committees discount hardest.
A 20-year PPA closes that gap. It matches the electricity contract to the useful life of the building and, more importantly, to the amortisation schedule of the debt. At 207 megawatts of contracted capacity, the annual electricity spend is large enough that a few cents per kilowatt-hour of price movement swings project economics materially; fixing or collaring that number is what allows a sponsor to model a debt service coverage ratio a lender will accept. This is the mechanism behind the headline: the deal that clears financing is not the one with the most megawatts, it is the one with the longest defensible price.
The people affected sit well beyond the two counterparties. Developers competing for the same capital are now benchmarked against this structure. Generators with dispatchable fleets gain a new class of long-duration offtaker. And prospective data center tenants — the hyperscalers and AI companies that ultimately lease the space — inherit whatever price the developer locked in, for two decades.
207 Megawatts Is a Campus, Not a Pilot
Capacity figures in this sector are easy to misread. A single enterprise server hall might draw a handful of megawatts. A 207-megawatt commitment is campus scale — the order of magnitude associated with multi-building developments serving large cloud or AI tenants, drawing power continuously rather than at a daytime peak.
Loads of that size cannot simply be requested from a utility on a normal timetable. Interconnection queues — the waiting lists grid operators maintain for new large connections — now run years in many US markets, and queue position has become a genuine asset. Contracting with an existing independent power producer that already owns generating plants is one of the few routes that does not begin with building something new. It does not eliminate the need for transmission and interconnection work, but it does change what the developer has to prove to get there.
What a 16% Move Prices In, and What It Doesn’t
A double-digit single-day move tells you something about relative scale: 207 megawatts is transformative for New Era in a way it is unlikely to be for Vistra’s consolidated results. Equity markets are reacting to the credibility conferred by the counterparty as much as to the contract’s economics, which were not disclosed.
It is worth stating plainly what a PPA is and is not. It is an obligation as well as an enabler. Depending on structure, the buyer may owe payment for contracted volumes whether or not the load materialises — the same term that protects the project against price risk creates exposure if the data center is delayed, downsized, or never leased. Until the price, the shape of delivery and the identity of the end customer are known, the market is pricing a durable input at an unknown cost against demand that has not been publicly contracted. That is a reasonable thing to be optimistic about; it is not yet a verified margin.
Background
Over the past two years, the constraint on building AI-capable data centers has migrated. First it was chips, then suitable land and buildings, then grid interconnection — the queue of applications that grid operators process before a large new load can be energised. The current binding constraint for many developers is securing firm, long-dated electricity at a price they can underwrite, which has pushed independent power producers into direct, multi-decade contracts with digital infrastructure companies.
Vistra is among the US generators that own dispatchable plants and sell their output into wholesale markets and to retail customers. For companies with that profile, contracting output to data centers trades merchant price upside for revenue visibility. On the other side, developers such as New Era Energy & Digital — which positions itself at the intersection of energy and digital infrastructure — treat a signed supply contract as a prerequisite for raising the capital to build.
IREN used to mine bitcoin. Now it rents out computing power for artificial intelligence — and Microsoft has signed on as a customer.
The company has $2.8 billion in new orders. That is money customers have agreed to pay over time, not cash sitting in the bank today.
Why an ordinary person should care: the buildings that run artificial intelligence cost enormous sums to build before they earn anything. A signed commitment from a buyer as large as Microsoft makes it far cheaper and easier to borrow the money to build them.
Moomoo reported that IREN Ltd has secured a contract with Microsoft and booked $2.8 billion in new orders, a combination the company presents as accelerating its transformation into an artificial-intelligence infrastructure provider.
IREN is a Nasdaq-listed operator that began as a bitcoin miner and has spent recent years redirecting its data center estate — land, substations and grid connections built for mining — toward hosting AI computing workloads. A named hyperscale customer and a multi-billion-dollar order figure are the two most material markers of that shift to date.
Executive Summary
Two facts carry this story: a dollar figure and a name. The $2.8 billion order book is the number that will travel; the Microsoft contract is the part that changes how the business is valued and financed.
The distinction matters because AI data center capacity gets built in one of two ways. It is built on spec, in the hope that rental rates for AI chips hold up long enough to repay the cost — or it is built against a signed, multi-year commitment from a buyer. The two look identical from the outside and cost radically different amounts of money to finance.
For the broader market, IREN is a test case for a pattern now visible across the sector: large cloud providers contracting directly with converted crypto-mining operators rather than waiting on their own construction pipelines. That arrangement transfers the demand risk off the operator’s balance sheet — and transfers a different set of risks, concentration chief among them, onto it.
The Counterparty Is the Number That Matters
The headline figure is $2.8 billion in new orders. The more consequential disclosure is the name attached to it. In AI infrastructure, capacity is either merchant — built speculatively and rented at whatever the spot market pays — or contracted, built against a signed offtake agreement. The physical racks are the same. The cost of capital behind them is not.
Here is the mechanism. A merchant fleet of AI accelerators is a commodity asset: its revenue depends on rental prices that move quickly, on hardware that depreciates fast, and on an upgrade cycle measured in quarters rather than decades. Lenders discount that cash flow heavily, so the build gets funded mostly with equity, which is the most expensive money there is. A multi-year contract with a large, creditworthy technology buyer converts the same racks into a stream of receivables, and lenders then price the risk substantially against the buyer’s credit rather than the operator’s. Nothing about the building changes; the interest rate does.
That is what makes a named hyperscaler contract structurally different from a strong quarter of rental demand, and it is why the signature is worth more analytically than the order total. The effect radiates outward. Operators with an anchor tenant can commit to chip allocations and power contracts earlier and more cheaply. Operators without one bid for the same accelerators and the same megawatts on worse terms. Chip vendors gain firmer order visibility; utilities and power developers gain load commitments they can actually plan generation and transmission around. The party squeezed is the marginal builder with capacity but no counterparty — the one whose economics depend on renting to whoever shows up.
What Transfers From Mining, and What Doesn’t
The reason converted miners are credible counterparties at all is that they already hold the genuinely scarce inputs. Bitcoin mining required cheap, abundant electricity at industrial scale, which meant acquiring land, building substations, negotiating power purchase agreements and — critically — securing a place in the interconnection queue, the multi-year waiting list to connect a large load to the electricity grid. Those assets cannot be bought quickly at any price, and they are precisely what an AI operator needs.
What does not transfer is the interior. Mining rigs tolerate heat, dust, single-feed power and interrupted service; a mining site that goes dark for an afternoon simply loses a day’s revenue. AI training runs do not tolerate any of that. They require redundant power paths, far higher rack power densities and usually liquid cooling, high-bandwidth low-latency networking between thousands of chips acting as one machine, and contractual uptime guarantees with financial penalties attached. Retrofitting a mining shed into that is substantial capital expenditure the mining business never had to carry.
The practical read, then, is that a converted site is a head start rather than a finished product — and that a hyperscaler signing a contract is meaningful third-party evidence that specific sites cleared that engineering bar. That is a stronger signal than any self-described transformation, because the buyer had every incentive to inspect before committing.
An Order Book Is a Promise, Not a Profit
An order book is contracted future revenue, and its meaning depends almost entirely on the term over which it is delivered. The same $2.8 billion is a very different business spread across three years than across ten, and the annualised figure implied by each is not something a reader can derive from the total alone. Orders also convert to revenue only as capacity is delivered, which means the recognition schedule follows a construction timeline, not a sales announcement.
The corresponding obligation is capital. Serving AI contracts means buying accelerators, energising power and installing cooling ahead of the revenue those contracts will eventually produce. In this business, a growing order book and rising cash consumption arrive together; the relevant question is the size and funding of the gap between them, not the size of the backlog.
Stated even-handedly: what is substantiated here is a named customer, a headline order figure and a clear strategic direction — all three are material and none are trivial. What is not yet substantiated is the margin profile, the contract term, the contracted capacity or the delivery schedule, and those four variables determine whether $2.8 billion is a transformative book or a demanding one.
Concentration Is the Price of Credibility
The anchor tenant that makes capacity financeable also concentrates the business. A contract large enough to move a company’s valuation is, by definition, a customer whose departure at renewal would move it just as far in the other direction — and large buyers know that when renewal terms are negotiated. Cheap capital today and pricing leverage against you later are two faces of the same agreement.
There is a second-order consideration worth stating plainly rather than dramatically. Hyperscalers lease third-party capacity alongside building their own, and the leased portion is a flexible variable in their planning: it can expand when their construction pipeline lags demand and contract when it catches up. That is a legitimate procurement strategy, not a criticism of any buyer. But it means operators in this position are underwriting capacity against a demand source that has an alternative and they do not. The mitigation is diversification of the order book, which is exactly why the split between the Microsoft contract and the remainder of the $2.8 billion is the most useful single disclosure still outstanding.
Background
IREN Ltd was founded in Australia as a bitcoin miner and listed on Nasdaq, building its business around securing large volumes of low-cost, largely renewable electricity and constructing its own data center sites to consume it. That model required the company to become, in practice, a power and land developer — acquiring sites, building substations and working through grid interconnection processes — with mining as the initial use of the capacity.
Since 2023, a broad reordering has pulled that asset base toward artificial intelligence. Demand for AI computing outran the supply of energised, grid-connected data center space faster than new construction could respond, while bitcoin mining economics tightened. Operators holding power capacity found that the same sites could host far higher-value workloads, and a number of listed miners repositioned as AI hosting providers. The competitive question across that group has been which of them can pair their power advantage with the engineering and the customer relationships that AI workloads actually require — which is why a contract with a named hyperscale buyer reads as a milestone rather than a routine sale.
