NIP Group Bought 4.47 EH/s With 247 Million Shares. Mining Is the Cover Story.

NIP Group didn't buy Bitcoin miners. It bought a second identity, and existing holders are footing the bill in shares.

That's the read on the October 20 announcement. The esports outfit you probably know from Ninjas in Pyjamas agreed to take over machines that are already humming at third-party hosting sites, roughly 4.47 EH/s worth. The consideration? Up to 247,091,740 Class A ordinary shares.

No cash. No published dollar figure. Just a quarter-billion-share slug and a hash rate number.

We regret to inform you that this is how a lot of mining M&A gets done in 2026, and it's worth understanding why before you cheer the hashrate.

The receipts

Stack the numbers up and a pattern shows.

This is the third tranche of an ongoing asset purchase. If it closes as described, NIP projects combined capacity of about 13.85 EH/s. Management models roughly 195 Bitcoin a month at full deployment. Do the arithmetic on that and the implied assumption gets interesting. Bitcoin is minting about 13,500 coins a month right now, so 195 BTC works out to roughly 1.4% of the network's block rewards. For 13.85 EH/s to earn 1.4% of the pie, the network hash rate has to sit somewhere near 960 EH/s.

That's the whole thesis compressed into one number. Their revenue projection bakes in a fixed slice of a network that grows every single month. Every new exahash that plugs in anywhere on earth quietly shaves that 195 down. Nobody's lying here. That's just what a model does. But it's a model, not a yield, and the source material practically begged people not to read it as one.

Then there's the share count. A quarter-billion new Class A shares is a lot of paper, and the disclosure doesn't attach a dollar value to the package. Which means the seller's payout floats with NIP's stock price until the deal closes. If the shares drift lower, the seller eats it. If they rip, the seller does great and existing holders get diluted into a smaller slice of a bigger machine.

But here's the part that rarely makes the headline. Already-deployed machines mean the capex clock is mostly done. No eighteen-month build. No transformer queue. No county board meeting about noise ordinances. That's genuinely valuable in a business where the gap between signing a power contract and earning a satoshi can run two years.

So the structure isn't stupid. It's just not free.

The AI bridge that isn't a bridge

Now the pitch. NIP frames computing infrastructure as connective tissue between entertainment, Bitcoin mining and future high-performance or AI workloads. Shared appetite for power, maintenance crews and big physical installs.

Fine. Sort of.

But let's be blunt about the hardware. A SHA-256 ASIC mines Bitcoin and nothing else. It's not a general compute chip. It's not going to run inference on a language model because both happen to live in a building with three-phase power. That silicon has one job until it dies. What actually transfers across these businesses is operational muscle, power access and the ability to keep thousands of hot boxes alive at 3am.

That's a real skill set. It's also not an AI business. Ask yourself: does anyone with a GPU cluster shortage call an esports company because it owns some S19s and a hosting contract? CT never misses. Except when it does.

And this matters because three completely different propositions keep getting mashed together in these announcements. Buying equipment is one thing. Turning that equipment into Bitcoin is another. Successfully pivoting into third-party compute services is a third thing entirely, and it's the hardest of the lot. Only the first one happened here.

Steelman the bulls

Credit where it's due, because there's a real case on the other side.

Preserving cash is smart when your core business is entertainment and your stock is a more liquid currency than your treasury. Issuing equity for hard assets when your shares are reasonably valued is cheaper than borrowing at 12% and pledging your miners as collateral. Ask the 2022 vintage how that collateral call felt.

Buying live hash also means revenue on day one instead of revenue in 2027. In a market where the halving keeps cutting the block subsidy, time in the chair is worth something. And an esports brand with global reach and existing infrastructure relationships isn't the dumbest parent for a mining operation. Ninjas in Pyjamas has been shipping physical events and digital products for years. Logistics isn't foreign to them.

If Bitcoin rips, 13.85 EH/s at full deployment is a lever. Higher prices lift dollar revenue even as difficulty climbs, and a company that funded the hardware with paper captures all of the upside without draining the balance sheet.

That's the bull case. It's coherent. It's just not the thing being sold.

My verdict

The hashrate is real. The AI angle is a press release. And dilution is the actual headline.

Nothing in this deal closes yet. It's an agreement, not a transfer. Terms can shift, the third tranche can slip, and the share count is capped at a number that assumes everything goes right. Watch the closing language and watch what the share count looks like on the other side. That tells you more about the health of this business than 195 BTC a month ever will.

Here's my position, and I'll commit to it. Mining is a commodity business with brutal margins, a difficulty ratchet that never sleeps, and power economics that decide everything. Adding 4.47 EH/s to an entertainment company doesn't change that math. It just adds a second set of problems to a balance sheet that already had some. The winners in a deal like this are the sellers, who trade depreciating silicon for equity in a listed company. The losers are the holders who find out what their percentage looks like twelve months from now.

That said, I'd rather see a company pay in shares for machines that are already running than in cash for machines that don't exist. There's a version of this that works. It's the version where mining stays a boring cash flow line and nobody pretends an ASIC is a GPU.

The timeline is undefeated. Another day, another saga. And somewhere, a shareholder is about to learn what 247 million shares actually costs.