Bitdeer’s latest numbers show a miner pushing hard on two fronts: more Bitcoin from its own rigs, and a bigger bet on AI infrastructure that may be far steadier than mining ever will be.
- Bitcoin output surged as self-mining capacity expanded fast.
- Revenue rose, but losses stayed wide, so profitability is still elusive.
- Norway is now a major AI/HPC bet with a long-term colocation contract.
- Debt and capex remain the real story behind the growth.
Bitdeer said it mined 2, 694 BTC in the second quarter of 2026, up from 565 BTC a year earlier, as its self-mining fleet expanded sharply. Revenue climbed to $228.8 million from $155.6 million in the same quarter last year, but the net loss also widened to $92.3 million from $62.9 million.
That is the classic mining trade-off: output growth looks great on a slide deck, but the balance sheet still has to survive the party.
The company said average self-mining hashrate rose 389% year over year to 69.5 exahashes per second, or EH/s. If you do not spend your days staring at mining spreadsheets, hashrate is the amount of computing power pointed at Bitcoin’s network. More EH/s usually means more production potential, although uptime, network difficulty, power costs, and hardware efficiency still decide whether that power turns into real money.
Bitdeer also said self-mining generated $168.4 million of quarterly revenue. June production reached 990 BTC, while May production came in at 921 BTC, about 370% higher than May 2025. The company ended the quarter with 150 BTC on its balance sheet, down from 1, 502 BTC a year earlier.
That drop came after Bitdeer said it sold its 943 BTC treasury in February for liquidity purposes. Management said that move did not signal a shift away from Bitcoin mining. Fair enough. Selling Bitcoin to keep the business funded is not the same thing as giving up on the business. It is still a reminder that even miners who love talking about “treasury strategy” need cash when the lights are on and the bills are due.
The broader financial picture is still heavy. In the first quarter of 2026, Bitdeer Technologies Group Q1 2026 Financial Results and posted a net loss of $159.5 million. As of March 31, the company reported $297.7 million in cash and restricted cash and about $1.9 billion in borrowings. That is not a sleepy balance sheet. It is a company leaning hard on leverage to fund growth.
And growth is now coming from more than Bitcoin mining.
On Aug. 4, Bitdeer disclosed a 16-year colocation and services agreement through its Tydal Data Center subsidiary with Volta Tydal AS in Norway. The deal covers 121 megawatts of critical IT capacity, supported by about 133 MW of total power. In plain English: Bitdeer is not just selling rack space. It is trying to become a serious infrastructure operator for AI and high-performance computing, or HPC.
The company said the agreement is expected to generate about $4.7 billion in payments during the initial term, with an eight-year renewal option that could lift the potential value to about $8 billion. Those are Bitdeer’s projections, not GAAP revenue or operating profit, and they should be treated that way. Gross contract value is not the same as cash collected, and infrastructure deals have a habit of looking tidier in presentations than they do in the real world.
Bitdeer said payments average roughly $202 per kilowatt each month during the first 16 years and rise 3% annually. Electricity costs will be reimbursed by the tenant. Management also projects average annual revenue of $2.4 million per IT MW and a project net operating income margin of about 90%. That margin figure sounds beautiful, maybe a little too beautiful, but it is still just a projection.
There are real safeguards and real caveats underneath the headline numbers. Volta’s obligations are expected to be backed by about $1.3 billion in letters of credit arranged by affiliates of J.P. Morgan and another global financial institution, subject to customary conditions. Bitdeer can terminate the agreement if Volta misses specified milestones. Volta also has a no-fee termination right after ten years.
That makes the deal meaningful, but not bulletproof. A 16-year contract sounds locked in until you remember that contracts live and die by financing, milestones, customer behavior, and boring legal fine print. The devil is always hiding in the appendices.
The Norway arrangement follows Bitdeer’s plan to convert the site into a 180 MW AI facility based on NVIDIA reference designs. In March, the company hired Data Center Installations AS for the conversion. Bitdeer also plans to develop two additional halls totaling 47 MW for potential AI and HPC customers in the second half of 2027.
This is part of a broader miner pivot that is getting harder to ignore. MARA Holdings has planned to acquire a Texas site capable of supporting up to 2 gigawatts of AI and digital infrastructure capacity. TeraWulf announced a 20-year data center lease with Anthropic that it said could generate roughly $19 billion over the contract period. Different companies, same message: pure Bitcoin mining is volatile as hell, and long-duration infrastructure contracts look a lot prettier to investors.
That logic makes sense. Bitcoin mining revenue swings with price, network difficulty, halving cycles, and power costs. AI and HPC hosting can offer longer-term contracted revenue and, if executed properly, more predictable cash flow. But “if executed properly” is doing a lot of work there. These are capital-hungry projects, and capital-hungry projects have a nasty habit of turning into expensive optimism when timelines slip or customers fail to deliver.
Bitdeer is trying to keep both engines running at once. It is scaling self-mining and SEALMINER hardware while also building a second growth track around data center capacity and AI infrastructure. That dual strategy could pay off. It could also turn into a costly juggling act if debt costs rise, construction drags on, or the AI side ends up less profitable than the pitch materials suggest.
The company’s Nevada move shows the same pattern. Bitdeer announced a $36 million Nevada factory for Bitcoin mining in Sparks, Nevada for SEALMINER Bitcoin mining machines, with commercial production scheduled to begin before the end of 2026. CEO Catherine Guo said the incentive package included reduced qualifying sales taxes. The point is clear enough: Bitdeer still sees hardware as core to its future, even while it talks up AI infrastructure as a second growth leg.
That matters strategically. Owning more of the mining stack can mean tighter margins, better supply control, and less dependence on third-party hardware vendors. But it also means more execution risk sits inside Bitdeer’s own house. If the factory slips, the Norwegian buildout gets messy, or debt becomes too expensive, the company will have more moving parts and more ways for things to go sideways.
Bitdeer shares rose about 1.5% in premarket trading Monday after the results, even though they had fallen about 15% over the previous month. That reaction makes sense. Investors like production growth and they like the AI angle, but they also know this is still a leveraged company with a lot to prove before it can be treated like a clean, durable infrastructure story.
Key takeaways
-
Why did Bitdeer’s Bitcoin output jump so much?
Because self-mining capacity expanded quickly. More rigs and more production and operations usually mean more Bitcoin, although network conditions and costs still matter. -
Is Bitdeer profitable yet?
No. Revenue rose, but losses also widened, which means the company is still in heavy expansion mode rather than harvest mode. -
Does the Norway deal guarantee billions in revenue?
No. Bitdeer’s $4.7 billion and $8 billion figures are projections, not guaranteed GAAP revenue, and the agreement includes milestone conditions and a no-fee termination right after 10 years. -
Why are Bitcoin miners moving into AI and HPC?
Because they already own power access, land, and data center expertise. AI hosting can offer longer-term contracted revenue than mining, but it also demands serious capital and disciplined execution. -
What is the biggest risk for Bitdeer right now?
Debt. The company already carries about $1.9 billion in borrowings and says it will use additional debt to fund more buildout, which means execution has to stay sharp.
Bitdeer is not pretending Bitcoin mining alone will carry the whole future. It is trying to turn mining cash flow, hardware manufacturing, and AI infrastructure into one bigger industrial machine.
That could work. It could also become a brutally expensive lesson in leverage if the AI side disappoints or the financing gets ugly. In crypto, as ever, the dream is easy to sell. The part where you actually make the numbers work is where the adults show up.
Further reading
For the AI-infrastructure angle and the miner pivot, these pieces add useful context:
