AIB Data Centers ( AIB): Power Play — AIB's Move from Bitcoin Hosting to AI Infrastructure

In this episode of the WTR Small-Cap Spotlight Podcast, Jolienne Halisky, CFO of AIB Data Centers Inc. (NYSE American: AIB), joins host Tim Gerdeman and WTR equity research analyst James Kisner. AIB is a power-first developer of AI and high-performance computing infrastructure. The company locks up executed utility power agreements before it breaks ground, then builds modular, liquid-cooled facilities that tenants lease and fill with their own GPUs. Halisky explains why power, not chips, is the real bottleneck in AI infrastructure, and how securing it first lets customers deploy compute months or years sooner. She covers the recent name change from BlockchAIn Digital Infrastructure, the shift from Bitcoin hosting to AI workloads, and how the build-to-suit model keeps hardware obsolescence off the balance sheet. Halisky also walks through the roughly 40 MW live today, a pipeline approaching 395 MW, the debt-free capital strategy, and the milestones investors should watch.

Transcript

SPEAKER_01 0:00

Welcome to the WTR Small Cap Spotlight Podcast. I'm your host, Tim Gerdeman, Vice Chair and Co-Founder and Chief Marketing Officer of Water Tower Research. Today I'm joined by Jolene Haliskey, CFO of AIB Data Centers Inc. NYSC American Ticker Symbol AIB. The company recently completed its corporate name change from blockchain digital infrastructure to AIB data centers in June. AIB Data Centers is a power-first developer of AI and high performance computing infrastructure. The company secures executed utility power agreements before breaking ground, then builds modular, liquid-cooled facilities leased to tenants who bring their own GPUs under TripleNet build the suit agreements. AIB operates roughly 40 megawatts of live cash-flowing capacity at a CLT 01 campus in the Charlotte market with additional capacity under development and a pipeline approaching 395 megawatts across the Carolinas, Minnesota, and Dallas Fort Worth area. The company began trading on NYSC American in March of 2026. Also joining is my WTR equity research colleague, James Kisner. Jolene and James, thanks for joining me today.

SPEAKER_00 1:19

Thanks, Tim. Good morning. It's an absolute pleasure to be here.

SPEAKER_01 1:23

So, uh, Jolene, to kick things off, for listeners new to the AIB story, can you give us a 30,000-foot overview of the company to include what does AIB do as a power-first AI data center developer? And what problems are you solving for customers today?

SPEAKER_00 1:40

Thanks, Tim. That's a great place to start, actually. The simplest way to describe AIB is that we are a power-first developer of AI and HPC infrastructure. So what that means practically is we go after the scarcest resource in our industry, that being utility power, before we do anything else. We lock up the executive power agreements with the utility companies, we secure the substations and the interconnects. And only then do we break ground. Once a site is energized, we build modular liquid cooled facilities, as you mentioned. We lease under long-term modified gross built to suit agreements with our tenants. Now, our tenants are neoclouds, AI native operators, HPC users, they bring their own GPUs. We deliver the physical infrastructure, that being the power, the cooling, the shell, the connectivity. The problem that we're solving is timing. The bottleneck in our AI infrastructure industry today isn't chips, it's power delivery. Interconnection cues are at most utilities measured now in years. By solving that first, we let our customers deploy their compute in months, sometimes years faster than what they otherwise could. So that kind of in a nutshell describes what we're doing and the problem that we're primarily solving.

SPEAKER_01 3:08

That's a great overview. I appreciate it. Very interesting space you're in, pretty hot. And with that, I'll turn it over to James to ask a few questions.

SPEAKER_02 3:16

Thanks, Tim, and welcome again, Jolene. So you know, as Tim mentioned in the intro, you recently moved from the blockchain digital infrastructure name to AIB data center. So what does that name change signal strategically?

SPEAKER_00 3:31

So the name change for us was really about clarity. The blockchain digital infrastructure name reflected our origins that we're very proud of. We've been operating a Bitcoin hosting data center since 2022, successfully operated that facility, but it wasn't where we were headed strategically. Our capital, our pipeline, and our leasing focus now are all pointed at AI and high performance computing. AIB data centers describes the business as it exists today and as we see it evolving. For investors, particularly institutional investors who are evaluating us against pure play digital infrastructure comparables, the new name makes it much easier to understand what they're buying and in what space we're operating in. The ticker stays AIB and the underlying business is the same, but the branding now matches our strategy.

SPEAKER_02 4:29

That makes sense. So um I want to kind of double-click a little bit on the power first aspect, um, you know, how you lock up utility power before you break ground. And I'm just wondering why you think perhaps power has become the real bottleneck for AI infrastructure, and how is going after that first set you apart?

SPEAKER_00 4:46

Great question. If you look at the constraints in AI infrastructure, that being chips, land, capital, and power, three of those four have gotten materially easier over the past 18 months. Chip supply, don't get me wrong, chip supply is still tight in specific SKUs, particularly, but nothing like it was that we saw in the last couple years. Land is available in most markets. Um we have a dedicated team of people that look for the right pieces of land across North America. And capital is flowing into the sector. As alluded, this is a hot sector right now. However, power remains the exception in many markets, and it's actually becoming more challenging, not easier to find those sites with that power procurement in place. So some markets are effectively closed to new large load customers in the foreseeable future as well, which adds additional complexity. And even where you can get an interconnection study done, the execution work to energization, coordinating with utility, the grid operator, and often state regulators can take years of sustained effort. So our approach is to do that work first before we commit construction capital. We first of all identify markets with available capacity. We build the relationship with the utility providers, we execute the power agreements, which is an expertise and a specialty unto itself. We secure the interconnect. That's being the last step. Once that's done, everything downstream, construction, leasing, financing becomes much more predictable. It also means when we approach a customer, we're not selling them the promise of some future power, CEO calls it fake power. We're selling them power that's already committed to the site.

SPEAKER_02 6:46

Yeah, that makes a ton of sense. We certainly heard about those uh those shortages uh of availability for power. So um, you know, I understand your model is owner agnostic, and so that means that tenants bring their own GPUs and you deliver the facility. Can you kind of walk us through how that works and why it lets you move quickly with less risk?

SPEAKER_00 7:05

Okay, sure. So you're absolutely right. Our our model is what's sometimes called powered shell plus. So we build up to white space, we deliver everything up to and including the cooling systems, and the tenant brings the compute, basically. It's structured as a modified gross lease built to suit the tenant. The tenant is responsible for their equipment, and utility costs are passed through. We deliver a facility built to their specifications and pick up the insurance, the real estate taxes, and any property management expenses. So there's a few reasons this works well for us. First of all, it takes the risk of technology off our balance sheet. Everyone knows GPU life cycles are short. What's state of the art today likely won't be state of the art in three years. That's almost guaranteed. By having tenants own the compute, we don't carry that obsolescence risk. Second, it lets us move faster. We don't need to procure GPUs. We don't need to build managed service capability. We focus on the piece of the value chain that we're best at, which is power and the physical infrastructure. And third, it aligns our incentives. Our tenants have significant capital in the building in the form of their own equipment, the GPUs, which reinforces their long-term commitment to the lease that we are building out the physical infrastructure for. For our customers, it means they retain control over their compute stack, which for neo-cloud and sophisticated AI operators, that's non-negotiable. Um, while getting to market much faster than they could have if they would have been engaged in building their own facility.

SPEAKER_02 8:58

Yeah, that's compelling. It seems like investors can get exposure potentially to this build that's happening without some of the uh corresponding risks you see in some of their models. Um you built a pipeline that runs well beyond your live capacity today across a few markets. How do you think about converting that pipeline into revenue?

SPEAKER_00 9:19

You know what, we think about it in stages. As you alluded to our pipeline, it's uh today comprised of a mix of sites in different phases, some with executive power agreements, some where we've secured land and are working through interconnection, some in earlier stage. So we don't think of pipeline as a single number that converts on a straight line basis. We think of it as a portfolio of options, basically, moving through each set of gates. So those gates being, first of all, site control, power procurement, permits, interconnection, tenant commitment, financing, construction, and energization. Each gate de-risks the next. And for us to even consider taking a site and control of a site, that power has to be available within 12 months. So meaning an ESA or a PPA is underway. And this is the, as I mentioned, the gates de-risk each other. The capacity we're most confident about is the capacity that's cleared the power and interconnect gates. That's what we point investors to as near-term revenue potential.

SPEAKER_02 10:38

That's a helpful explanation. So the business, as you talked about earlier, has shifted from Bitcoin hosting toward AI and high performance computing workloads, and you've broadened the customer base along the way. So, what's driving that shift and what does it mean for the durability of your revenue?

SPEAKER_00 10:54

Great question. You're asking all the tough ones this morning. Here's a few things. Uh, let me let me gather my thoughts on that one. So a few things are happening in parallel. The Bitcoin hosting business, that's our real business, and we and I spoke of that earlier. But it has characteristics that are difficult for public equity investors to remain attracted to. The revenue is highly correlated to a commodity price, that being Bitcoin. Customer credit profiles vary widely, and contract durations are much shorter. AI and HPC is fundamentally comprised and builds a different profile. Customers are large, well-capitalized enterprises, typically, which is very attractive for investors. Contracts are typically structured with a 10-year initial term with two to three renewal options. Pricing is negotiated based on power delivery and facility economics, not tied to a spot commodity. So it brings some stability to the cost structure. And the demand tailwinds, driven by AI training and inference buildout, are more secular rather than cyclical. For durability of revenue, the AI HPC mix means, first of all, longer average contract life, secondly, strong counterparty credit, and lastly, the cash flows that look much more like traditional infrastructure. So that's what makes the business financial at scale.

SPEAKER_02 12:31

Yeah, that makes sense. I don't think anybody can argue that uh about any of those points you made about AI customers uh versus crypto. Um, you know, you since you're the CFO, why don't we talk a little bit about some uh financial topics? Um, you have an ambitious growth plan. Takes a lot of capital to do that. So how are you thinking about funding that growth while keeping the balance sheet in good shape?

SPEAKER_00 12:54

So that our balance sheet right now is you know no debt, and we're very protective of that. We, you know, we have two principles that guide how we think about capital. First of all, we we match the tenor of the capital to the tenor of the asset. And I mentioned before, leases are structured uh into 10-year initial terms with the renewals. So we want long-dated capital structures to match against those cash flows. Secondly, as a I was just referencing how pleased or how comfortable we are with our nice clean balance sheet with that is debt-free. We keep that corporate balance sheet flexible enough so that we're not forced to raise on someone else's timeline. We can raise on our timeline and evaluate the best alternatives without feeling pressured. Also, we use a mix of instruments. At the corporate level, we've done a follow-on equity offering most recently. We completed that earlier this year. At the project level, we layer in construction and term debt at that SPV level. And we partner with potentially LP private equity type partners on structured pieces where they fit the risk profile. Over time, we would expect a significant portion of our project capital to come from non-recourse project financing tied to contracted lease revenue, which is in all reality a much more efficient way to fund infrastructure rather than issuing corporate equity dollar for dollar. So we've been very thoughtful about staging. We want to fund each project as it de-risks. We talked about the gates of de-risking earlier, rather than going ahead and pre-funding capacity that hasn't been leased yet. And that protects our shareholders from dilution and ultimately projects that might not move forward.

SPEAKER_02 14:59

Okay, that sounds like a very thoughtful approach. So um let's just turn to profitability for a second. Um, you know, energy costs is something talk people talk a lot about in data centers. I'm just kind of wondering how that impacts your business. Do you have to manage it? And how should investors just broadly think about margins as you grow in AI and HPC?

SPEAKER_00 15:19

Great question. And it's a complicated one, so I'll try and structure my response accordingly to capture all the key points. Uh okay, energy is the single largest operating cost in the data center, as everyone knows. So, you know, really are focused on how we manage that and how that defines our margin profile. So the two pieces that matter most is what we contract that power for, the the price of that power contract. And number two is how efficiently we use it. So on the contract side, we lock in those utility rates through our power agreements before we sign a tenant lease, and our lease structures pass fully pass through energy costs through to the tenant. So we're not carrying commodity price risk on our PL. We're earning our economics on the physical infrastructure, not arbitraging power prices. On the efficiency side, our facilities are designed for liquid cooling right from the start, which is materially more efficient than legacy air-cooled type designs that we've experienced in the past. That better PUE translates directly into lower delivered cost per kilowatt hour for our tenants, which supports pricing power over time. We can't model our margins in against one factor in isolation. This is a capital-intensive business, and the right way to evaluate returns is on unlevered project IRR and stabilized yield on cost basis. That's how we run it internally, and that's the framework we point analysts towards as well.

SPEAKER_02 17:02

Great. That's helpful clarification. So you mentioned this before that that your lease terms are typically 10 years with renewal options. Just how do those contract links and structures shape your revenue and margin visibility? And you know, what are the risks to those long-term contracts and how do you manage that?

SPEAKER_00 17:19

Okay. So even though you know, this we feel this space is much more de-risked than our previous operating space in the data mining sector, you know, when you look at these leases, the 10-year initial term with renewal options is the market standard for the built-to-suit data center leases. And it's the structure our tenants want as much as we do. For us, it gives us contracted cash flow visibility that supports project financing and predictable long-term returns. For the tenant, it gives them certainty of tenure in a facility they've customized for their workload and made a significant investment in on their end. On the risks, you're absolutely right. The main ones are counterparty credit, technology evolution, and renewal. So on the credit piece specifically, we underwrite our counterparties very carefully. Our anchor leases are with only high-quality, well-capitalized operators. And we look hard at balance sheet strength, cash flow, and their strategic commitment to the site before we be we before we even sign. And that's both for the tenant and if there's an off-taker in place at the at the time that we're negotiating the lease. On technology, our facility design is built for liquid cooling and high density workloads, which is where the market is going. And because tenants own the equipment, technology refresh happens at their expense and on their timeline. It doesn't force us to refit or retrofit the facility, but the facility. So campus level unit economics improve meaningfully as we build out capacity within a site.

SPEAKER_02 21:19

All right. Sounds like there could be a lot of operating leverage potential in there from various places. Um, as we wrap up, um, you know, Julian, what are the key milestones you'd point investors to over the next few quarters that would tell them uh that you're executing on the plan?

SPEAKER_00 21:33

Okay, so uh probably this one is a little bit easier to answer because I would direct investors to the same places that we are watching and monitoring so very closely on a day-to-day basis. So, first of all, is our execution at the CLT01 site. That's our current operating campus where Bitcoin data mining operations existed that are being decommissioned, and the um additional capacity beyond the 40 megawatts that's live today, another 25 megawatts is under development. So investors need to watch for the construction and energization milestones that are embedded within that project. Secondly, additional lease commitments, as every new lease that's signed converts pipeline into capacity and into contracted revenue. And those are the events that most meaningfully change our forward earnings profile. We actually just had uh two analysts initiate coverage on us, and it was very interesting to see how their profiles aligned and correlated with our expectations as well. Third, advancement of the next sites in our pipeline. I think I mentioned before we have a team of people dedicated to site procurement across North America. So moving those sites through the gates that I described earlier, from power to permits to construction start, we'll communicate as those milestones are reached for each site because those are very exciting milestones for us. And fourth, on the capital side, we'll continue to be thoughtful about how we fund growth. Investors should expect to see us match capital raises to project milestones rather than raising ahead of need. Finally, we've been building out our investor relations function, being just newly public. We're still developing how we how we interact with our investors. So listeners can expect to hear more from us, definitely, through channels like this one, through conferences, through more frequent shareholder communications, including something we're very excited about and want to highlight is we're we're actually having our ring the bell at the New York Stock Exchange this Friday, July 10th. So building durable relationships with our investor base is a key priority for me as CFO.

SPEAKER_02 23:59

Great, very exciting stuff. So uh thanks again, Jolene, for coming, and I'll pass it back to Tim.

SPEAKER_01 24:04

Thanks, James, and uh Jolene, thanks so much for joining us today and walking us through the very interesting AIB data centers uh investment theme. Uh have a great day and thanks for your time.

SPEAKER_00 24:14

Thank you, Tim and James. Thanks for having me join you today. It's been a pleasure.

SPEAKER_01 24:21

Thank you for listening, and don't forget to subscribe, as well as visiting www watertowerresearch.com to stay up to speed on the company's small cap written research reports, podcasts, fireside chats, industry specific symposiums, and conference schedules. We will see you next time for another edition of the WTR Small Cap Spotlight Podcast. Finally, a special thanks to the producer and editor of the podcast, Krista Fitzpatrick.