Core Scientific Is Quitting Bitcoin Mining. Mining Is Now Just 7.5% of Revenue

Seven and a half percent. That's all Bitcoin mining is worth to Core Scientific now.

One of the biggest publicly traded miners in North America says it'll stop mining Bitcoin by the end of the year and show up in 2027 as a pure-play data center company. Mining, the thing that gave the company its name and its ticker and its whole identity, brought in just 7.5% of quarterly revenue. The other 92.5% came from renting out power and floor space to AI compute tenants.

Think of it this way: a company spent four years building a brand around SHA-256, and the math quietly voted for someone else's workload.

From Chapter 11 to landlord

The arc here's worth sitting with for a second. Core Scientific went through Chapter 11 in late 2022, when Bitcoin was trading near $16,000 and the cost of electricity made every machine in the fleet a money loser. It came out in January 2024 with a leaner balance sheet and a strange new idea: instead of spending capital on ASICs, sell the space, the cooling, and the grid interconnection to companies that needed compute for something other than hashing.

That bet turned into hosting contracts with CoreWeave. The numbers moved fast. Deals that started in the low billions grew, then grew again, and by 2025 the contract value was being discussed in the double-digit billions. Core Scientific wasn't selling hashrate anymore. It was selling megawatts.

Now the company wants to finish what those contracts started. By year-end, mining goes dark. In 2027, the pitch deck says pure-play data center provider. No hedging, no "we'll keep some machines running just in case." Just a straight handoff.

And honestly, why would management keep the miners? A megawatt pointed at Bitcoin produces whatever hashprice allows on any given Tuesday. A megawatt under a multi-year AI hosting contract produces revenue you can put in a spreadsheet and show a lender. For everyday users, nothing changes overnight. For the equity story, everything changes.

Follow the megawatts

Here's the part most people get wrong about miners. They were never really in the Bitcoin business. They were in the power arbitrage business, wearing a Bitcoin costume.

Bitcoin mining rewards you for two things: cheap electricity and access to capital for hardware. The hashing itself is commoditized. Any operator with the same power price and the same machines earns the same margin. There's no moat in the computation. There's only a moat in the interconnection queue, the substation, the water rights, the county permit.

Which is why the AI wave didn't just lift miner stocks. It repriced the one thing miners actually owned. Hyperscalers are desperate for powered land near fiber, and the queue to get a new grid connection in most US markets runs three to five years. A miner sitting on 500 megawatts of already-energized capacity isn't a mining company anymore. It's a landlord with an option on the best real estate in tech.

So, who wins? CoreWeave and its peers, clearly. They get capacity without waiting in line. Core Scientific shareholders get contract revenue that doesn't swing 40% on a hashprice chart. Lenders get to underwrite something with visibility past next quarter. And the local utilities get a customer with a load profile that's flatter than a mining farm's, which matters more than people admit.

Who loses? This is where I'll take a harder line. Bitcoin loses a sophisticated, publicly traded operator with a real balance sheet. Every miner that leaves is one fewer institution with the capital to defend the network through a bad cycle. Hashrate doesn't collapse, it never does, but it does migrate. Toward private operators. Toward places with loose disclosure and cheaper power and less scrutiny.

Is that a security problem? Maybe not today. But here's a question worth asking: if the most profitable use of a brand-new mining facility is to never plug in a miner, what exactly is the long-term plan for the network's security budget after the next two halvings?

Nobody wants to price that. The block reward went from 6.25 to 3.125 BTC in April 2024, and the fee market hasn't come close to picking up the slack outside of a handful of inscription spikes. Miners are supposed to be the last line of defense for finality, and we're currently paying them with an asset whose issuance curve is designed to shrink. The pivot to AI is a rational response to that. It's also a signal.

Here's why the plumbing matters. Core Scientific isn't exiting because Bitcoin failed. It's exiting because AI pays better for the same electron, and because a data center lease has something Bitcoin mining has never had: a contract. That's not a betrayal of the industry. That's the industry telling you what it's actually worth.

The mining trade is over. The power trade isn't.

Strip out the branding and what you've is a company that figured out it was sitting on the scarcer asset all along. Compute is abundant relative to demand right now. Interconnection isn't. Power isn't. Land next to a substation isn't.

My read: expect more of this, and not in slow motion. Watch what happens to the mid-tier public miners with decent grids and aging ASIC fleets over the next two or three quarters. Some will announce their own hosting pivots. Some will get acquired for their megawatts and nothing else. A few will insist they're staying pure Bitcoin, and those are the ones I'd want to see a balance sheet for.

Bitcoin will keep producing blocks every ten minutes regardless. That's the whole point of the design.

But the next time someone tells you mining is a growth industry, ask them what percentage of their revenue comes from hashing. If the answer is under 10%, they aren't a mining company. They're a real estate play that used to be one.