Mining Under Pressure: Why CleanSpark, BitFuFu, and Canaan Cut Output in June

June brought the mining industry an unpleasant surprise. Three public companies at once — the American CleanSpark and the Chinese BitFuFu and Canaan — reported a decline in production of the first cryptocurrency over the past month. And it happened at a time when network difficulty had fallen to its lowest level since the start of the year: formally, conditions for mining were about as comfortable as they had been all half-year. Let us break down what happened at each company, how the economics of mining works, and why the industry is increasingly looking toward artificial intelligence.

What the June Reports Showed

The numbers speak for themselves. CleanSpark produced 614 BTC in June against 671 BTC a month earlier. BitFuFu's result fell from 177 coins to 125, and Canaan dropped from 90 to 64. All three companies trade on stock exchanges, so they disclose operating metrics monthly — their reports have long served the market as an early indicator of the whole industry's health.

The picture is worrying: the decline was recorded by companies of different sizes, with different geographies and different approaches to production. This is no longer an isolated glitch at a single site but a symptom of broader processes affecting all cryptocurrency mining.

There is another thing worth understanding: the monthly reporting of public companies is the industry's most up-to-date snapshot. Financial results come out once a quarter, while production data arrives every month, which is why analysts use it to judge where mining is heading overall.

Who Public Miners Are and Why Their Reports Matter

All three companies are public: their shares trade on an exchange, and they are obliged to report to investors regularly. That is exactly why the market sees their results almost in real time, whereas private players stay in the shadows and disclose very little about themselves.

This transparency makes public miners a convenient barometer for the entire industry. They account for a noticeable share of the network's total computing power, and their behavior — buying equipment, selling reserves, changing strategy — usually reflects what is happening at hundreds of non-public players, for whom cryptocurrency production has also become the core business.

There is a flip side. Being public forces companies to think constantly about their share price: investors expect metrics to grow every month, and a weak report immediately hits the stock, even if the cause of the dip is temporary and purely technical. Hence management's drive to explain any shortfall in advance and in as much detail as possible.

CleanSpark: Minus Three Exahashes

For the American company, the reason for the decline is entirely technical. CleanSpark's average operating hashrate fell in June from 46 EH/s to 43 EH/s — roughly 6.5%. For the company this is a direct loss of its slice of the overall pie: the less computing power it has, the less often it finds blocks and the fewer coins it receives.

CleanSpark's reserves, meanwhile, remain impressive — 13,924 BTC on the balance sheet. The company is still one of the most prominent players in the US industry, and its monthly reporting traditionally sets the tone for expectations across the sector.

CleanSpark itself tied the drop in production specifically to the decline in average hashrate, without pointing to external circumstances. For the industry, such honesty is telling: when capacity falls, no market conditions can rescue the result.

BitFuFu: Leases Gone, Own Capacity Growing

BitFuFu's story is more interesting. The company's total computing power fell from 19.5 EH/s to 15 EH/s — almost a quarter. The reason is a reduction in leased capacity: part of the company's mining ran not on its own equipment but on someone else's, and dropping the lease was immediately reflected in production.

At the same time, BitFuFu is building up its own mining infrastructure. In June the company brought 1,200 Antminer S21 XP units online, pushing its own hashrate up to 3.5 EH/s. In July it plans to connect another 2,000 devices.

So in the short term production sagged, but structurally the company's mining is becoming more self-sufficient. Leasing someone else's capacity is a quick way to scale volumes, but an expensive one: it eats into margins and makes the business hostage to someone else's rates. Owned equipment takes longer to pay off, but it gives control over the cost of mining.

Canaan: The Power Grid Let It Down

Canaan's reason is even simpler: scheduled maintenance of the power grid at one of its sites. The company confirmed that these works were exactly what limited its mining in June and caused production to fall from 90 to 64 coins.

This is a good illustration of how dependent mining is on infrastructure. Even a brief drop in the electricity supply immediately hits the month's result: the equipment sits idle, while competitors keep on with cryptocurrency production and take the freed-up share.

It is worth noting that Canaan is known primarily as a manufacturer of mining equipment, and its own production is a secondary line of business. That is why its figures in absolute terms are more modest than CleanSpark's, but the trend is telling all the same.

June's Paradox: Difficulty at a Low, Yet Production Falls

The most curious part of this story is the context. In June, network difficulty fell to its lowest level since the start of the year. Logically this should have made mining easier: the lower the difficulty, the fewer computations are needed to find a block and the more coins go to those still in the game.

But all three companies still ended up down on production. That means the issue is not general market conditions but internal factors: hashrate, leases, and energy. This is an important nuance — it shows that even favorable difficulty will not rescue mining if a company has problems with its own capacity.

What is more, the drop in difficulty is itself an alarming signal. It means that some equipment around the world has disconnected from mining, and the network has adjusted to the reduced power.

How Mining Works

To make sense of these reports, it helps to recall how mining works. It is a process in which specialized computers run through billions of hash variants, trying to find a suitable one and close out the next block of transactions. Whoever manages it first receives the network's reward and the fees — and that is how cryptocurrency production happens.

There is no magic here: mining is a lottery in which the chance of winning is directly proportional to the computing power invested. That is exactly why large companies chase hashrate, and why mining has turned into an industrial sector with data centers, long-term electricity contracts, and megawatt substations.

The economics of mining, meanwhile, comes down to just two components: how many coins you managed to get and how much electricity it took. Everything else is derived from these two numbers, which is precisely why any drop in power shows up so quickly in the reporting.

Hashrate: Mining's Key Metric

Hashrate is the total speed of computation. It is measured in hashes per second, and at industrial scale we are talking about exahashes (EH/s) — quintillions of operations every second. For mining, hashrate is roughly what engine power is for a car.

When CleanSpark loses 3 EH/s and BitFuFu 4.5 EH/s, it means literally the following: their share of the network's overall mining has shrunk, and they receive fewer coins. No managerial tricks can compensate for this — mining is too straightforward in its mathematics.

That is exactly why companies publish hashrate alongside production: it shows what their mining is really capable of the following month. The equipment works around the clock, and any dip in average power converts straight into coins not earned.

Network Difficulty and Its Effect on Production

Difficulty is a built-in regulator. The network automatically recalculates it roughly every two weeks so that blocks are found on average every ten minutes, regardless of how much equipment is connected to mining.

When some miners leave, difficulty falls — and those remaining have it easier. That is exactly what happened in June. But a drop in difficulty only partly compensates for the loss of one's own hashrate: if a company has cut its own capacity, its mining will sag anyway.

The result is a peculiar balance: difficulty protects those who stay in mining, but it does not repeal the laws of physics and economics. Either you spend electricity and get coins, or you switch off the equipment and cede your share to others.

Cost of Production: When Mining Stops Paying Off

Behind the individual reasons for the decline lies a shared economic problem. According to JPMorgan analysts, the full cost of mining a single bitcoin is currently around $78,000, while the market price of bitcoin hovers near $62,500. A gap of nearly 20% means a significant part of the industry is producing coins at a loss.

The bank notes that the price has held below the cost of production for five months running, and roughly one miner in five is now in the red. In such a situation, cutting mining is not weakness but a rational decision: it is better to pause part of the equipment than to burn electricity for coins worth less than they cost to make.

MARA is a vivid example. The company increased its computing capacity by 25% compared with the fourth quarter of last year, bringing it to 66.4 EH/s, yet production volumes fell by 15%. The cost of the electricity needed to produce a single coin rose over the year from roughly $31,000 to $48,000. It is electricity, not equipment, that today decides who stays in the game.

CoinShares offers similar estimates: the average cost of mining for public companies is approaching $80,000 per coin. At current quotes, this makes the old business model frankly unprofitable and forces companies to look for alternatives.

February 2026: How the Industry Got Here

The June decline is far from the first alarming signal this year. In early February, the mining revenue indicator fell to a record low, and companies began switching off equipment en masse. Back then CleanSpark's stock fell 10% in a day, MARA 11%, TeraWulf 8.5%, and Riot 4.8%.

CleanSpark executive Harry Sudock called that downturn historic and pointed to two causes: the collapse in the price of the first cryptocurrency and winter storms in the US. Bad weather in Texas and Tennessee triggered a sharp spike in electricity rates, and the energy bill is the main expense line for any site.

Newhedge analysts at the time forecast a difficulty drop of more than 13% at the next recalculation — the largest fall in the metric since China banned the crypto industry in 2021. The June reports are a direct continuation of that story: the industry still has not returned to its former pace.

Record Sales of Reserves

The pressure on mining economics is also visible in how companies behave. According to JPMorgan, citing TheEnergyMag, in the first quarter of 2026 alone MARA, CleanSpark, Riot Platforms, Cango, Core Scientific, and Bitdeer collectively sold more than 32,000 BTC to fund operations.

That is more than those same companies sold in all of 2025, and a new quarterly record: the previous high of 20,000 BTC was set in the second quarter of 2022, during the crash that followed the Terra-Luna collapse. Total mining reserves today are estimated at roughly 1.8 million coins, against 1.86 million at the end of 2023.

The logic here is simple. When mining does not pay off, accumulated reserves become the only source of real cash. Companies spent years piling up coins as a strategic asset, and now they are selling them off to pay for electricity and service their debts.

The Pivot Toward Artificial Intelligence

This is precisely why the mining industry is pivoting en masse toward AI. The data centers, substations, and electricity contracts built for cryptocurrency mining are also an excellent fit for high-performance computing — but the pay is far more stable.

The scale is impressive: the industry has already signed more than $70 billion in infrastructure contracts. Among the largest are Core Scientific's $10.2 billion agreement with CoreWeave, TeraWulf's $12.8 billion projects, and Hut 8's $7 billion contract. To finance the transition, companies are borrowing: IREN raised $3.7 billion and TeraWulf $5.7 billion.

For many participants, mining is gradually turning from the core business into a flexible background load switched on when spare capacity is available. This fundamentally changes the industry: revenue from AI computing depends far less on the coin's price, and investors value that predictability more highly than a bet on mining.

CleanSpark and a New Business Model

CleanSpark itself is a telling example. The company's CEO, Matt Schultz, stated outright that the new agreement marks a transition from bitcoin mining to a diversified digital infrastructure model and will allow the company to begin commercializing its energy assets.

This is not a one-off decision: back in October 2025, CleanSpark announced it was developing a data center line of business and planned to adapt its energy sites for artificial intelligence loads. The new lease simply rounds out that profile.

Other mining players are taking a similar path. Bitdeer sold off its coins entirely and is raising funds for data centers, Core Scientific is retooling for AI, and Bit Digital intends to stop production altogether and focus on other directions.

How the Stocks Reacted

The market took the news in different ways. CleanSpark's shares rose to $13 (+5%) and BitFuFu's to $1.42 (+7%), while Canaan slipped to $0.20 (−1.5%).

Investors' logic is clear: a drop in production frightens them less by itself than the outlook for the business. Where mining is complemented by energy and AI infrastructure, the market is willing to forgive a weak month. Where a company remains a pure mining play, any dip is taken more painfully.

The market also reacts not only to the numbers themselves but to the tone of the commentary. When a company explains a decline with scheduled works on the power grid, investors take it calmly. But when the cause is lost capacity or a dropped lease, the reaction is noticeably more nervous — because that is a question of strategy, not the calendar.

Another example is telling: Hive Digital Technologies reported a net loss of $148.4 million despite growing revenue and a doubling of production volumes. More coins does not mean more profit — that is the main lesson of the current cycle for all of mining.

What Is Happening to the Network

The mass outflow of capacity has already affected the network. Bitcoin's hashrate fell from roughly 1,160 EH/s in 2025 to around 920 EH/s, and for the first time in several years a run of difficulty decreases has been recorded.

For the ordinary user this is barely noticeable: transactions go through, blocks are found, fees have not spiked. But for the industry it is a structural shift — mining has ceased to be an unconditionally growing business and has become one of several ways to use energy infrastructure.

It is worth remembering that the network does not become more or less reliable because of this. The protocol was designed from the outset for participants to come and go: difficulty adjusts automatically, and blocks keep being found at the same rhythm. The shareholders of specific companies may suffer, but not the users who are simply sending transfers.

There is a flip side here too. The fewer independent participants remain in mining, the higher the concentration of capacity among the largest players. For now this is not critical for the Bitcoin network, but the trend is worth watching closely.

Bottom Line: What to Expect Next

The June reports from CleanSpark, BitFuFu, and Canaan are not a catastrophe but an honest snapshot of the industry's condition. Each company had its own reason: hashrate, leases, the power grid. But the backdrop is the same — mining today is balancing on the edge of profitability, and companies are optimizing everything that can be optimized.

What comes next depends on the price. If bitcoin returns above the cost of production, the switched-off equipment will be turned back on, and mining will quickly restore its volumes. If not, the pivot toward AI will accelerate, and cryptocurrency production will become, once and for all, a supporting rather than a primary revenue line for public companies.

For investors the conclusion is likewise mixed. A weak month at an individual company is not yet a diagnosis; it matters more to look at the structure of the business: does it have its own energy, what are its rates, how heavy is its debt load, and does it have contracts outside of production. It is these factors, rather than one-off output figures, that determine who survives the current cycle.

That is precisely why monthly reports should be read not as a verdict but as an indicator. They show that mining has stopped being a simple story of "buy the equipment and collect the coins" and has turned into a complex business at the intersection of energy, finance, and computing.