Exclusive | BitDeer: From Bitcoin Miner to "AI Landlord"

Author | Lin Wanwan

Editor | Sleepy.txt

No one initially thought that the real bottleneck of AI is not capital, not large models, but electricity.

With large-scale training running at full capacity for extended periods and AI inference operating 24/7 non-stop, a problem arose: there wasn't enough electricity, and chips were forced to sit idle. The U.S. has relatively lagged behind in electricity grid infrastructure in the past decade, with new significant loads connected to the grid taking 2–4 years, making "readily available electricity" a scarce commodity across the industry.

Generative AI has brought to the forefront a raw and harsh reality: it's not the model that's lacking, it's the electricity.

As a result, the story took a turn, and crypto mining companies, the group that initially treated electricity as a "means of production," began moving from the periphery to the center stage of capital.

Iris Energy (IREN) is a prime example of this path. This year, IREN's stock price surged by nearly 600% at one point within the year, reaching from $5.12 to $75.73 in a 52-week period. Sensing the continued allure of Bitcoin's price surge, it decisively allocated power to transform its self-built AI data center.

When giants like Microsoft wielded $9.7 billion worth of long-term orders, the market for the first time intuitively understood the realistic path of "from mining to AI," where electricity and land precede GPUs and customers.

However, not all mining companies, like IREN, chose to bet everything on AI. In this electricity-driven hash power migration, there is also a steady force worth noting — Bitdeer.

Bitdeer Technologies Group (NASDAQ: BTDR), founded by crypto legend Wu Jihan and headquartered in Singapore, strategically holds nearly 3GW of global power resources, steering clear of the shallow pitfall of relying on others for "power supply" from the start. As the AI wave swept over, Bitdeer did not opt for IREN's aggressive "All-in" approach but retained profitable Bitcoin mining as its "core business," while prudently upgrading some mining facilities into AI data centers.

This "able to attack, able to defend" strategy has made Bitdeer the best example to observe how global players are contemplating and positioning themselves in this hash power race.

To delve deeper, we interviewed Wang Wenguang, Vice President of Global Data Center Business at Bitdeer, hoping to shed light on the current global AI electricity shortage and their perspective on whether mining companies transitioning to AI data centers is driven by capital speculation or a genuine need for AI. In this series of questions, we engaged in an in-depth conversation.

Why Is the Power Shortage in the United States So Severe?

Insightful Inquiry: Let's start with a fundamental question, do you think electricity prices will continue to rise in the future?

BitDeer: I believe they will, as it is a crucial aspect of future supply and demand dynamics.

Insightful Inquiry: Regarding the power shortage in the United States, there is a belief in the market that it is difficult to obtain an "electricity license" in the U.S. What is your take on that?

BitDeer: It's not that this so-called "electricity license" is hard to obtain, but rather that the physical expansion of the power grid is not keeping pace. In the years following the relocation of heavy industry outside the U.S., the U.S. power grid infrastructure did not see systematic expansion. When mining companies moved into the U.S. in 2021, a significant amount of "already grid-connected and PPA-signed" electricity was locked up by these companies. With the influence of ChatGPT, pure AI players entered the scene and only then did they realize the abundance of immediately usable power in mining farms.

This explains why major corporations are willing to collaborate with mining companies. Instead of waiting 2-4 years to build up 500MW from scratch, they prefer to spend 12 months renovating existing facilities.

Insightful Inquiry: When did the industry truly realize that "inference also consumes a lot of power"?

BitDeer: It was probably after the widespread adoption of GPT-4. As companies embedded models in customer service, office work, search engines, risk management, and other areas, the long-term and scenario-based nature of inference demand meant that power consumption did not decrease as initially envisioned.

This brought about two types of changes.

One is engineering upgrades: from stronger air cooling to liquid cooling/hybrid heat dissipation, cabinet power, power distribution paths, fire protection, and monitoring have all been elevated to the level of AI data center operations.

The other is resource strategy: electricity has truly become the primary bottleneck. People are no longer just talking about "buying GPUs" but are instead focusing on acquiring electricity and grid access upfront, securing long-term PPAs, grid connection schedules, cross-regional capacity backups, and, when necessary, obtaining power upstream like mining companies (self-generation/direct procurement).

In fact, in the mining industry, we have long seen the same trend. Chips can scale infinitely (silicon comes from sand), but power expansion is not as flexible. We have used natural gas self-generation for powering mining farms in Canada, following the same logic. Today's AI is almost identical.

Insightful Inquiry: How does the power usage of AI data centers differ from traditional internet data centers?

Bit Deer: It's not about incremental change, but about orders of magnitude change. In the past, a 20-30 MW traditional internet data center was considered large, but today AI data centers have power demands of 500MW or even 1GW. AI has transformed data centers from a "rack business" to a "power engineering" business, where everything needs to be reevaluated: power lines, substations, cooling, fire protection, redundancy, PUE... The experience from traditional internet data centers is still useful but no longer sufficient.

Dongcha: Why has "electricity" become the scarcest upstream resource?

Bit Deer: Chips can be scaled because they come from silicon and capacity management; however, electricity is difficult to scale because it comes from power generation and grid upgrades. In the past, the mining industry tried to "source energy upstream," including undertaking self-power generation projects in Canada. The path AI is taking is similar—whoever controls the power first will have the advantage in deployment time.

AI New Battlefield: From "Grabbing GPUs" to "Grabbing the Power Grid"

Dongcha: When mining companies transition to AI data centers, what specific changes are needed? In the past, people said, "Bitcoin mining power can be used to run AI," but mining chips (ASIC) and GPUs needed for AI are not compatible. So why can mining companies now "provide AI computing power?"

Bit Deer: The global mining industry used to be divided into two camps—Bitcoin relied on mining chips ASIC for high efficiency but single-purpose use, while Ethereum relied on NVIDIA GPUs for their general-purpose use, but has exited the mining stage after transitioning to PoS.

Therefore, what is referred to as "mining farms transitioning to AI" in the market today almost always refers to Bitcoin mining farms transforming. The key point is that mining farms are no longer "hashing" but are upgrading themselves into AI data centers.

This involves infrastructure upgrades—removing ASIC racks and installing GPU servers; upgrading the "just enough" power systems to professional-grade power distribution with N+1/2N redundancy; upgrading traditional air cooling to a cooling system capable of supporting high-density GPUs; and standardizing and making facilities such as room sealing, dust prevention, and fire protection auditable.

Completing these four steps transforms a crypto mining farm from a "mining workshop" to an "AI data center."

Why can mining companies self-build faster than AI giants? Electricity.

AI is a business of "electricity and heat," and the construction time frame for AI data centers is 3-4 years, with time being the biggest barrier to entry. Mining companies happen to hold these "hard assets," giving them a head start in the transition process.

Insight: In the past few days, Microsoft and Amazon have successively signed multi-year AI contracts with cryptocurrency mining companies. Iris Energy (IREN) signed with Microsoft for a total value of 9.7 billion, spanning 5 years; another company, Cipher, signed with Amazon Web Services for 5.5 billion over 15 years. This is seen as the first batch of cases of collaboration between mining farms and tech giants. What is your take on this?

BitDeer: Iris Energy is a forward-thinking Australian company that has been mining in the United States for a long time.

Iris Energy's shift towards AI is like a signal flare. At a time when the price of Bitcoin is high and peers are still expanding their mining operations, it diverted some of its power to invest in a self-built AI data center. Subsequently, AI companies took the initiative to approach them.

The real tipping point came from the Hyperscalers' substantial commitments—such as Microsoft's around 9.7 billion dollars. For the first time, the market clearly saw that between mining companies and tech giants, it's not just about "technology integration," but about the "exchange of power and time."

The hype around AI has magnified the infrastructure needs, opening up collaboration opportunities.

Insight: Why are top mining firms more likely to be chosen by American AI tech giants at this stage?

BitDeer: Because of "available power + engineering delivery speed." The site selection and grid connection from the previous cycle of mining firms have now become the upfront capital for AI data centers. Time is the biggest discount factor, directly determining who can go live within the window, acquire customers, and generate rolling cash flow.

Insight: Is the land selection requirement for AI data centers challenging?

BitDeer: Overall, not really. In the United States and most countries, the real scarcity is power, not land.

The reason is simple: places that can receive large amounts of power are mostly energy-rich areas (natural gas fields, coal mining belts, near hydroelectric power plants), with sparse population and low land costs.

For example, Bitdeer's large data centers in Norway and Bhutan are located far from population centers, where power resources are concentrated and land costs are low. In the United States, similarly, these industrial parks are not in the city's core area but in more remote locations where land is easy to find and inexpensive. The "first principle" of site selection is electricity and grid connection, and land usually follows power, not being the primary bottleneck.

Insight: AI is now being referred to as the upstream business of "iron, electricity, and land," even likened to another form of real estate. What is your take on this?

BitDeer: After the emergence of large models, the energy consumption of AI far exceeds most people's expectations.

Initially, everyone thought "training consumes power, but inference is lightweight." However, in reality, after inference became mainstream, it also exhibited long-term high power consumption. As ChatGPT and DeepSeek became more prevalent and more terminals were connected, the baseline power consumption of inference continued to rise.

From an engineering perspective, AI is fundamentally a resource-intensive industry:

· Chip side: During training, acceleration cards are running at nearly 100% load, naturally consuming high power;

· Data center side: The heat density is much higher than that of traditional servers, leading to a significantly higher PUE, and the cooling itself also consumes a large amount of electricity;

· Scale side: The power demand of AI data centers has leaped from 20–30MW in traditional internet data centers to 500MW, or even 1GW, a level that was almost unimaginable in the era of traditional internet data centers.

Therefore, likening it to "real estate" only captures half of the picture. It does indeed require land, buildings, and a long cycle (construction cycles often take 3–4 years), but its life and death are determined by electricity and heat, whether it can get access to high-capacity power on time, implement N+1/2N redundancy, and efficient cooling. In this respect, its strong dependency is very similar to that of iron, electricity, and land.

What are the characteristics of AI data centers?

Insight: What are the unique characteristics of data center construction in the United States?

BitDeer: Due to power constraints and historical paths in the United States, Hyperscalers often need to personally intervene and cooperate with mining companies to obtain available power.

Insight: Is it possible for foreign companies to establish AI data centers in the United States?

BitDeer: In simple terms, AI data centers are a strongly regionally focused business. When it comes to landing with power capacities in the hundreds of megawatts and thousands of calories, it is still primarily the domestic giants in the United States leading the way. We are only discussing AI data centers here, excluding traditional internet data centers.

Insight：Will AI Data Centers evolve into geopolitical tools? Will this affect your decision-making?

BitDeer: I agree with this assessment.

At the core of AI lies data, which is inherently subject to sovereignty and security constraints. To prevent data leakage and security risks, regions around the world are tightening relevant policies: even though the U.S. allows foreign investment to build data centers, as AI gains access to more and more data, most countries are likely to move towards "on-premises deployment, local compliance, and data localization."

In simple terms, U.S. AI stays in the U.S., Middle Eastern AI in the Middle East, European AI in Europe, and regionalization will be a long-term trend.

Industry Landscape and Potential

Insight: Apart from IREN and BitDeer, who in the mining industry has more potential to transition to AI data centers?

BitDeer: To see who stands a chance, first see if they have access to high electricity capacity, then see if they can quickly convert mining farms into GPU data centers. The ones with grid connection, land, substation, N+1/2N redundancy, liquid cooling/high density are the most likely to attract AI projects.

For the other type that is purely hosting/light-asset, without control over electricity and industrial parks, transitioning to AI data centers will be passive.

In the U.S., companies like Riot, CleanSpark, Core Scientific, TeraWulf, Cipher that have resources in hand and reliable expansion capabilities are more likely to be targeted by tech giants.

So, the conclusion is straightforward: electricity is the ticket, transformation capability is the speed; only when both are in place can you lead the race.

Overall, it depends on who holds "high-quality, sustainable large-load available electricity." Companies with more self-owned grid-connected resources have more potential; the model that relies on hosting and lacks self-owned energy and industrial parks is not advantageous in this round of structural transformation.

What Is BitDeer Thinking?

Insight: What is BitDeer's strategy and path for "mining to AI"?

BitDeer: Mr. Wu Jihan's strategy has always been to cover the entire industry chain. BitDeer holds around 3GW of power and industrial park resources, which is our biggest underlying advantage.

When we first entered the AI field, we did not anticipate that "electricity" would become a core bottleneck, so we initially took a self-built, self-operated approach: we partnered with NVIDIA to become an NVIDIA PCSP, deployed a small-scale H100 cluster in Singapore, launched our own AI Cloud, and provided external training services. This project has been successfully implemented.

Subsequently, we also established a second data center in Malaysia. As the Hyperscalers entered this arena and began collaborating with mining companies, we simultaneously upgraded high-load campuses to AI data centers: we have announced the overall transformation of a 180MW site in Norway into an AI DC and the conversion of a 13MW site in Washington state, USA.

Fundamentally, the essence of AI is very similar to Crypto mining—both are businesses based on "electricity + infrastructure." We have end-to-end capabilities in power supply, campuses, and computing operations, so transitioning to AI has been relatively smooth for us.

Insight: What are the core differences between BitDeer and other mining companies like IREN?

BitDeer: Three points. First, we will not fully transform into an AI enterprise; based on calculations, the current stage of Crypto Mining is still more profitable than AI data centers, and the mining industry offers stable cash flow and better returns.

Our second advantage is the international engineering organizational capability. The BitDeer team's engineering organization and execution capabilities are unparalleled worldwide. For the same AI data center, the common rhythm in the United States is two years, but we typically can achieve it in one and a half years. This is achieved through parallel advancement and supply chain coordination, synchronizing key aspects such as civil engineering, electromechanical, power distribution, and cooling to compress the usual 24-month cycle to about 18 months, thereby creating usable capacity faster.

The third is the company's strategy of maintaining stability: the AI industry is very young, even younger than Crypto, so we do not go "all-in" and instead pursue a longer-term development pace.

Insight: Currently, where is BitDeer's power infrastructure mainly located?

BitDeer: BitDeer's power infrastructure is currently mainly distributed worldwide with approximately 3 GW of power and related infrastructure, covering five countries: the United States, Canada, Norway, Ethiopia, and Bhutan. This infrastructure supports the construction and operation of mining and AI data centers.

Cost and Financing

Insight: I saw a Goldman Sachs report mentioning that an AI data center might cost around 12 billion US dollars. Is it really that expensive?

BitDeer: Indeed, it's huge, with the scale being "tens of times." Here's an easily understandable comparison in "layman's terms": Building 1 MW for a Bitcoin mining farm (in the US) costs approximately 350,000 to 400,000 US dollars. But building 1 MW for an AI data center costs about 11 million US dollars. This is because the investment in an AI data center is a compound of "heavy mechanical and electrical work + heavy standardization," along with grid connection queueing, environmental assessment/energy assessment, regional compliance, with a typical timeframe of 18 to 36 months.

You will find that AI data centers are fundamentally not about "buying a few more cards," but about connecting a piece of land to form a "city of electricity" that can consume 500 MW to 1 GW, ensuring proper electrical connectivity, heat dissipation, redundancy, compliance, all of which are very costly.

Insight: Where does the money come from? Do you need financing?

BitDeer: Honestly, everyone needs financing.

Let me share, here are some common financing methods in the industry:

1. Project Financing/Infrastructure Loan: Pledge the industrial park + equipment, rely on long-term leases or hash rate offtake (customer commitments to buy your hash rate for many years) to reassure the banks.

2. Equipment Leasing/Leaseback: Lease out GPUs and some electromechanical components, spreading the cost over a long period, so you don't have to pay a large sum of cash upfront.

3. Long-term PPA (Power Purchase Agreement): First lock in the electricity price and available capacity, and then the debt side would be willing to offer a low-interest rate.

4. Tie-up with major companies: Large customers/major companies provide minimum consumption guarantees, prepayments, guarantees, or even joint ventures (JVs), allowing you to obtain cheaper funding.

In collaborations between IREN, CoreWeave, and Google/Microsoft, you can see these terms reflected.

Insight: Will BitDeer also need financing? Will an announcement on partnering with major companies be made soon?

BitDeer: I can't say much publicly about this right now.

Conclusion

Not long after the interview ended, BitDeer revealed its next steps in the capital market.

On November 13, BitDeer announced that it would raise $400 million through the issuance of convertible senior notes, granting the initial purchasers an option to purchase additional notes of up to $60 million within 13 days, with a maximum fundraising scale of $460 million. The new funds will be used for data center expansion, ASIC miner development, AI and HPC cloud business expansion, and general corporate purposes.

In a time where electricity has become the scarcest upstream resource in the AI industry, where this $460 million will ultimately land and how many megawatts of new load it picks up will largely determine BitDeer's position in the next round of computing power competition.

For BitDeer, this money is more like putting the judgment from the interview into the balance sheet: one end connected to the cash flow bedrock of the mining industry, the other end connected to the long and snow-covered business line of AI data centers. It may not immediately reflect in the revenue and profit of the next quarterly report, but it will slowly reshape the power structure of the computing power business in the coming years—who is eligible to sit at the negotiation table, and who can only queue up on the grid connection list waiting for electricity.

Looking ahead from the results, this round of AI infrastructure story is not that complicated: electricity has truly become the upstream, time has become the new currency, and the industrial parks and grid connection indicators in the hands of mining companies have become "old assets" that others can't buy even if they have the money.

As the noise about models and applications slowly fades, the market will most likely go through the ledger again: whose narrative is most resonant will no longer be important. The company that can connect every megawatt of electricity in a power shortage world and run steadily is the one qualified to stay at the next stage of the table.