Who owns the stack: From Bitcoin to AI, the race for power is going off the grid.

This article first appeared in The Energy Mag. The original article can be viewed here. Energy Mag (formerly Mining Mag) provides news, information and insights on the energy-calculation-market nexus.

In the first part of this series, we explored a basic idea: Bitcoin mining It was never just about digital currency. It is designed to be a long-term energy system that operates on a supply schedule lasting more than a century.

In the second part, we explored how this system is not unique. Bitcoin. Modern AI data centers are built on the same physical foundation—chips, power, cooling, and infrastructure—all of which work together to turn electricity into electricity. Bitcoin mining And AI processing at scale.

The AI ​​boom has dramatically increased demand, requiring critical infrastructure and high-efficiency technology to support grid resilience, ultimately reshaping how the digital future is built.

Across the technology spectrum, developers are moving beyond traditional grid reliance, directly saving electricity through discrete energy assets. This “bring your own power” approach to AI may feel new, but it's a playbook Bitcoin Miners have been searching for years.

This type of vertical integration has become a defining characteristic of the brand. Bitcoin mining Industry. As companies begin to expand into new markets, including the United States, they strive to isolate existing infrastructure. However, this wave of expansion did not stop at data centers – operators directly owned and partnered with energy companies to improve the reliability, sustainability and affordability of the energy supply.

This third section builds on evolution. if so Bitcoin mining And AI data centers share the same system, the next question is how companies handle themselves. What is visible is a spectrum of business models – from outsourcing to full vertical integration.

The more capital a company controls, the broader and more it affects its costs, performance, and bottom line returns.

But it is worth noting that this stack is not permanent – and neither are the companies that work in it.

as a Bitcoin mining And AI data centers measure on the same basic infrastructure, the line between their business models begins to blur. The two seemingly disparate industries are becoming a common system, with stakeholders moving in real-time.

The entry point: asset-light deployment

At the basic level, participate in both Bitcoin mining And AI computing starts with deploying hardware.

in Bitcoin miningThis means taking ASIC machines and inserting them into devices made by others. Bitcoin. In AI computing, the peer is deploying GPUs into data centers, where they are used to train models and run workloads for customers.

In both cases, companies own the machines – but not the infrastructure.

That infrastructure is provided by colocation operators, which run power, cooling and compute physical space at scale. Historically, this has been seen as a supportive function. It is becoming one of the most important parts of the business.

Ink is no longer just about hosting machines – it's about creating power and infrastructure.

On the other side Bitcoin miningThis model has been included in the industry for a long time. Companies like ABTC deploy hosted miners through parent infrastructure operators like Hut 8 (NASDAQ: HUT ), while others like Cango operate fleets hosted by Bitmain-run facilities. In each case, the separation between hardware ownership and infrastructure operation defines the business model.

That same structure is now emerging in AI.

Companies like Fluidstack are partnering with infrastructure providers including Cipher and TeraWulf to deploy GPU clusters. BitDear (NASDAQ: BTDR ) is actively developing its AI detection capabilities, including a proposed 180MW facility in Tidal, Norway, which is currently in lease negotiations.

As demand for AI accelerates and power becomes the limiting factor, the cost of existing grid access infrastructure is increasing. Many sites are built first. Bitcoin mining They are now well-positioned to support AI workloads, and over time, an increasing share of these facilities—especially across the US and Europe—are expected to transition to AI and high-performance computing applications.

In this sense, coloring is not the only entry point into the stack.

It is becoming a bridge between two industries – computing power, infrastructure and demand in one and growing system.

Controlling the infrastructure

As companies move up the stack, the next step is owning the physical environment.

At this stage, companies don't just deploy hardware. Instead of relying on third-party hosting, operators build or acquire their own facilities, including data centers, distribution centers and cooling systems.

This change significantly changes the work. Infrastructure ownership allows operators to control energy costs, increase performance and reduce dependence on external suppliers.

But the ever-increasing value of infrastructure is not only in the buildings – it is in the power connections associated with them.

That flexibility is playing out in industrial assets that were once considered obsolete, allowing companies to turn underutilized resources into powerful engines for growth.

Companies like Alcoa have begun exploring the sale of idle aluminum smelter sites to digital asset firms like NYDIG, and Century Aluminum sold a Kentucky facility to TeraWolf, which is now focused on AI and high-performance computing.

Many of these sites are at risk of closure due to the gradual exodus of high-paying industrial jobs to other countries. But they share one important characteristic: they are already connected to the power grid by measurement.

That connectivity—often the hardest and slowest part of building new infrastructure—suddenly became a valuable asset in its own right.

As a result, infrastructures reimagined to support facilities originally built for heavy industry are getting a second life. Bitcoin mining And AI workloads in the same way. This is bringing critical technical roles to the US, rebuilding key infrastructure and accelerating the nation to global leadership in technology and innovation.

In this environment, owning infrastructure is not just about controlling operations. It is about ensuring access to energy systems that can handle increasing demand while supporting overall resilience.

Bring your own power

But even that pool of grid-connected infrastructure is limited.

The number of industrial sites with high potential linkages is limited, and most have already been identified or rezoned by major industries. As consumer demand for computing accelerates—especially with AI—the amount of power required will require solutions to maintain grid resilience by deploying new technology solutions.

In other words, the limit is not limited to infrastructure. The energy grid is self-sustaining. That pressure is now forcing a broader shift.

Operators in major energy markets are facing a new reality: connecting large loads to the grid is becoming more complex and uncertain. Therefore, regulators are beginning to revisit how large energy users are integrated into the system.

In regions such as PJM and ERCOT, grid operators are beginning to adapt their frameworks to respond to demand from data centers and other high-load users. New rules and regulations are emerging to regulate how large data centers connect to the network, how costs are allocated, and how reliability is maintained amid rapidly growing demand.

To address these challenges, more and more operators are going beyond the grid altogether.

A clear example of this approach is the partnership between Amazon and Talen Energy to develop data center infrastructure with nuclear power generation capacity. While AWS doesn't own the energy assets, the structure actually balances the bill with a specific energy supply — demonstrating the same principle that's been in place for a long time. Bitcoin mining Works.

in Bitcoin miningThis includes living with energy sources that have not been used for a long time. Companies like New West Data flare Gas From the oil fields and use that energy to stimulate Bitcoin Miners for additional cash flow. In the year In 2020, Grange Generation became the first so-called power generation to participate directly Bitcoin miningrehabilitating assets that were closed due to competition in the energy market.

In AI computing, a similar model is emerging. Data center developers are working with or directly building power generation assets, including natural gas GasNuclear and above all, renewable energy.

This “bring your own power” model transforms electricity from a cost center to a strategic asset. It allows operators to verify pricing, ensure availability, and align compute capacity with power delivery.

in Bitcoin miningThis evolution is only visible and growing over the years.

A clear example is Bitfarms. The company has historically operated as an auto-mining business, owning the infrastructure and deploying its own computing power. But by acquiring Stronghold, Bitfarms has moved upstream into power generation, gaining direct control over power assets. Later, it was transformed into Kebele infrastructure, which represents a wide transition away from pure Bitcoin mining Towards a Model for Understanding AI and High-Performance Computing.

In practice, Bitcoin Companies are building knowledge to fuel the demand base for future technologies, including AI.

Full vertical integration

For some operators, even taking power ownership is not the last step.

At the high end, companies can control almost all aspects of the computer system, including power generation, infrastructure, hardware deployment, and chip design.

With AI computing, hyperscalers (such as Amazon Web Services, Microsoft, and Google) are starting to move in the same direction—developing custom chips, ensuring long-term power supplies, and building large-scale data center campuses tailored to their workloads. in Bitcoin miningThis model is no longer theoretical. It's already taking shape.

Canaan first Bitcoin ASIC designers with the Avalon brand, have gone beyond hardware to run their own mining infrastructure. In recent years, it has ramped up its proprietary computing power by deploying its own machines at locations it controls directly or through partnerships. Earlier this year, Canaan further strengthened its strategy by acquiring 49% of Cipher Digital. share In Texas Joint Ventures with WindHK, a wind electric power plant, moving its exposure up in the stack.

A similar direction can be seen with Bitdeer. Initially focused on cloud mining and proprietary operations, the company has steadily expanded its control over infrastructure and increased its dedicated computing power to around 70 EH/s. It has also moved into power generation, including acquiring land and licensing a 101MW plant in Canada.

At the same time, Bitdeer is expanding horizontally into AI processing. The company is exploring high-performance computing colocation opportunities with tenants and has begun deploying its own GPU infrastructure for AI cloud services.

This dual expansion—stack to power and AI workloads—shows how the boundaries between Bitcoin mining and data center infrastructure are beginning to dissolve. At this stage, the objective is not only efficiency. It is to reach.

By leveraging each layer of the stack, operators can maximize end-to-end performance, reduce exposure to external constraints, and define their own capacity limits.

Although few companies today fully occupy this position, the direction of travel is clear. Nearer operators will move towards full integration, shifting more from key energy users and developers of digital infrastructure.

Same stack, different places

What emerges from this comparison is not the history of two separate industries, but a common system with multiple points of engagement.

Bitcoin mining and AI data centers differ in their workloads and clients. But at the structural level, they work through the same ownership process, from asset-light distribution to infrastructure ownership, to direct energy storage, and finally to full vertical integration.

After all, those places are not fixed. Companies constantly reposition themselves—moving up the stack to gain control or to capture new sources of consumer demand. These dynamics reinforce the Bitcoin-AI coexistence approach: Secure high-volume power contracts and instantly monetize proprietary Bitcoin mining power while seamlessly reconfiguring computing infrastructure for high-margin AI.

Bitcoin miners, importantly, started solving these problems early, and AI companies are now reaching the same conclusion. The main difference is not the system itself, but how each company chooses to navigate.

In the next section, we'll take this one step further: how these models are starting to converge—and what they mean for the future of energy, computing, and capital.

This article first appeared in The Energy Mag. The original article can be viewed here. Energy Mag (formerly Mining Mag) provides news, information and insights on the energy-calculation-market nexus.