The press release was polished. The timeline was tight. The promises were loud.
But the only number that matters now is $41.9 million. That’s the penalty Core Scientific paid to walk away from Block’s Proto mining chips. Let that sink in. A company chose to burn $41.9 million in cash rather than take delivery of Jack Dorsey's 3nm ASICs.
This isn't a 'pivot.' This isn't a 'strategic realignment.' This is a funeral. And the body is Block's ambition to challenge Bitmain and MicroBT.
Context: The Great Miner Migration
For those who haven't been watching the tape, the narrative around Bitcoin miners has shifted faster than a flash crash. The old story was simple: secure the network, earn the block reward, hodl the Bitcoin. The new story is about AIDC — AI Data Centers.
Core Scientific, once the poster child of institutional mining, is leading this charge. Their Q4 2024 and Q1 2025 reports show a company that has moved from a pure-play Bitcoin miner to a hybrid infrastructure provider. They signed a 15-year contract with AMD to lease out their power capacity. The estimated revenue from that deal? $14 billion.
Against this backdrop, the Block chip deal happened. In 2024, Core signed a contract to take delivery of 15 Exahash worth of Proto chips — Block’s 3nm mining ASICs. This was supposed to be Jack Dorsey’s revenge on the Chinese duopoly. A validation of the 'open, decentralized mining' ethos.
It lasted less than 12 months.
Core Analysis: The Order Flow Tells the Truth
Let’s cut through the marketing. The order flow is the only truth.
Core Scientific terminated the contract. They paid $41.9 million to walk. This isn't a 'delayed delivery' or a 'mutual parting of ways.' This is a financial penalty for breach of contract. In the world of institutional finance, walking away with a nine-figure penalty is a screaming signal. It means the expected value of taking delivery was negative. It means the chips were not worth the price.
The industry standard is the Antminer S19 series and the Whatsminer M50 series. Bitmain controls roughly 70-80% of the market. MicroBT holds another 15-25%. These machines have proven reliability, established supply chains, and known efficiency ratios (J/TH).
Block’s Proto chip was an unknown variable. And here’s the dirty secret of the ASIC market: it’s a winner-takes-most game. You cannot enter this market by being 'good enough.' You have to be significantly better on price or performance to displace the incumbents. Core Scientific looked at the numbers on the Proto chip, compared it to the Antminers they could buy on the open market, and did the math.
$41.9 million is a tax on a bad decision. It’s cheaper than running a fleet of underperforming hashers for 3 years. It’s a smarter trade.
The Block team will tout a 'healthy pipeline of demand.' But the reality is: their only identifiable major client walked. That pipeline is a ghost. When the first institutional test ends in a penalty, it’s not a 'setback.' It’s a veto.
Contrarian: The Blind Spot on the Board
Here’s the counter-intuitive take that most analysis misses: this is not just a win for Core Scientific. It’s a warning sign for the entire Bitcoin mining ecosystem.
Everyone is cheering Core’s pivot to AI. 'Smart move,' they say. 'Capturing value,' they claim. I say: watch your leverage.
Core Scientific’s balance sheet is deep, but their debt pile was legendary. They emerged from Chapter 11 in 2024. They still have a thousand points of failure. The AMD contract is a 15-year estimated revenue of $14 billion. That’s an estimate. It is not a guaranteed P&L. If the AI capex cycle rolls over — if Meta, Microsoft, or Google cut their orders from AMD — Core will be left holding the bag for hundreds of megawatts of power capacity, just like they were left holding Block’s chips.
The market is pricing in a smooth transition from 'Bitcoin mining' to 'AI infrastructure.' I see a rougher path. The order flow is shifting, but the liquidity is thin. The next 12 months will show who really knows how to run a data center.
Takeaway: The Levels to Watch
The takeaway is brutal but simple.
For Block (SQ): The divergence is clear. The payment business (Cash App) is a cash cow, but management is burning capital on vanity projects — Tidal, TBD, Web5, and now this chip failure. Jack Dorsey’s 'Bitcoin-first' vision is costing shareholders billions. Watch the Q2 earnings. If they don't announce a complete write-off of the Proto business, the sell button is your friend.
For Core Scientific (CORZ): The transformation is real, but the runway is long. The AMD contract is a multi-year catalyst. The risk is execution and debt refinancing. If they can convert that 15-year estimate into real revenue without a liquidity crunch, they'll be the standard. If they hit a speed bump, the leverage will burn them again.
For the rest of us: Watch the S19 and M50 secondary market prices. If Block dumps their inventory of unsold 3nm chips at a discount, it will pressure the entire margin structure for mining hardware. That’s when the smart money buys in.
Mentorship is scarce; self-education is mandatory. Learn the order flow, not the press releases.
Liquidity dries up when everyone is looking away. Right now, everyone is looking at the AI pivot. I’m looking at the penalty. That’s where the signal lives.