On February 20, 2026, Bitcoin mining difficulty jumped 14.73% in a single adjustment.

With the same mining rigs and the same power costs, revenue fell by nearly 15% overnight.

Many mining rigs were already hovering around the break-even line. The latest difficulty spike made things worse.

Mining companies, however, are not making the same choices. Bitdeer sold all of its Bitcoin and turned to bet on AI computing power.

BitFuFu, another Nasdaq-listed miner, took the opposite path: its February output held up against the trend at 227 BTC, while it kept 1,830 BTC and $40 million in cash on its balance sheet without selling a single coin.

The Singapore-based Asian mining major has more than 640,000 users, with businesses spanning cloud mining, self-mining and mining rig hosting.

CEO Lu Liang came out of Bitmain, the same system that produced Bitdeer founder Jihan Wu.

Two companies from the same lineage are facing the same crisis: one is selling coins and pivoting; the other is calmly accumulating Bitcoin.

The difference lies in their revenue mix.

One BitFuFu business contributed 83% of its February output while quietly shifting away the risk of mining losses.

That business is cloud mining.

What Was Cut Was Not Computing Power, but Loss-Making Computing Power

BitFuFu’s total computing power fell 10.8% in February, but not because mining rigs broke down or customers left.

The company said it proactively cut a batch of long-term, low-margin hosting contracts.

With Bitcoin mining difficulty surging this year, the industry faces a hard reality: older, power-hungry mining rigs, such as the S19 series, lose money every day they run once electricity costs exceed $0.05 per kWh.

So BitFuFu chose the cleaner route: it removed these unprofitable “power hogs” from its hosting list, kept only efficient new machines, and stopped letting outdated equipment burn electricity for no return.

The effect was clear.

Its managed power capacity fell to 463 MW, meaning the overall footprint did shrink, but the remaining machines reached an average efficiency of 17.5 J/TH, a mid-to-high level for the industry.

More importantly, daily Bitcoin output rose from 7.4 BTC to 8.1 BTC. The operation got leaner, but production got stronger.

But cutting machines and saving power alone cannot make output rise as capacity falls.

BitFuFu’s ability to lift production while shrinking its balance sheet depends on another line of business working in the background.

Why Others Cannot Easily Replicate Cloud Mining

Cloud mining is BitFuFu’s core card.

In February 2026, the company produced 227 BTC, with nearly 80% coming from cloud mining. In the third quarter of 2025, cloud mining generated $123 million in revenue, about 70% of total revenue.

Put simply, this is a landlord-style model with locked-in economics: customers buy computing power in advance and pay service fees, while BitFuFu delivers mining output under contract.

The risk of Bitcoin price swings is effectively borne by the buyer. BitFuFu earns fixed service fees and gets paid in good markets and bad.

But cloud mining has long had a terrible reputation in crypto, largely because too many operators disappeared with customer funds. In February 2025, the U.S. Department of Justice said the two founders of cloud mining platform HashFlare had pleaded guilty in a $577 million case; its actual mining capacity was less than 1% of what it claimed, making it essentially a Ponzi scheme.

Earlier, Genesis Mining, once the world’s largest cloud mining platform, effectively ground to a halt after stopping contract sales in 2021, and users’ funds remain frozen.

In December 2025, security research firm Brave New Coin even said the probability of fraud among cloud mining platforms was close to 100%.

BitFuFu has managed to stand in this wreckage because of three things.

First, it has real mining rigs. It is one of Bitmain’s few global S-level customers, giving it priority access to new machines, and it recently signed large purchase orders for the latest mining rigs.

It has actual hardware running, rather than just a website taking payments.

Second, its technology can keep pace with scale. Its self-developed Aladdin system manages large fleets of mining rigs and can slice computing power precisely for different customers, while keeping availability at a high level.

That operational capability is something most makeshift teams simply do not have.

Third, compliance costs provide a backstop. As a Nasdaq-listed company, BitFuFu is regulated by the SEC and publicly discloses its financial data, making the cost of fraud far higher than for unlisted platforms.

Compliance, however, does not mean there is no controversy. BitFuFu’s revenue is highly concentrated: in the first half of 2025, its three largest customers contributed 58% of total revenue, and its largest single customer accounted for 31%.

If major customers choose not to renew, nearly one-third of the company’s revenue could disappear directly.

At the same time, the SEC has not yet clearly defined cloud mining contracts, and the regulatory stance could tighten at any time.

BitFuFu uses the cloud mining model to keep Bitcoin price volatility outside the door. What it leaves for itself is another kind of fragility: excessive reliance on a small number of major customers, with its lifeline partly in other people’s hands.

The Balance Sheet Is Strong, but for How Long?

BitFuFu’s February moves were clear: cut computing power and squeeze out profit.

In a challenging market, that conservative approach is beginning to show results.

Start with the balance sheet. As of the end of February, the company held about $40 million in cash and 1,830 BTC, giving it a meaningful cushion.

The debt side is even more notable.

Media outlet CCN estimated that mining companies raised more than $4.6 billion in 2025 through debt issuance and share offerings, pushing leverage higher across the sector. BitFuFu moved the other way: its loan balance fell from $40 million last September to $15 million, a drop of more than 60%. While others borrowed to expand, it repaid debt and shrank.

Its path has also diverged from sister company Bitdeer. Bitdeer cleared out its Bitcoin holdings and bet on AI, making a bold move. BitFuFu has not been in such a hurry. It has not shouted slogans about an AI pivot, but it still holds 463 MW of power capacity and self-owned mining farms in the U.S. and Africa. The industry view is that this infrastructure could be switched to AI compute leasing at any time, a case of keeping the option without saying much about it.

For now, BitFuFu’s financial position looks healthy.

Low debt, cash, Bitcoin holdings and power reserves make it unusually steady among mining companies.

How long that steadiness lasts depends on Bitcoin’s price. If the price falls, payback periods for cloud mining contracts lengthen, and fewer customers will be willing to buy.

Cloud mining is BitFuFu’s core revenue source. If that segment starts to shrink, the company’s cash-flow engine will slow.

The February data proves one thing: BitFuFu’s defensive posture is working.

It cut capacity without hurting output, and reduced leverage without damaging cash flow.

It stabilized its base by locking in cash flow upfront through a prepaid cloud computing power model, while using self-mining as a floor.

That low-leverage, countercyclical mix is hard to find in the sector.

But defense is still defense.

The key question in March is not whether BitFuFu can continue to “hold on,” but when and how it shifts to offense: will it keep accumulating Bitcoin and wait for the market, or open new monetization channels for its power and mining farm resources?

The answer will come in its next monthly report.