BTCB2: The BIP 110 Fork Came Back With a New Engine. What Its BLAKE2b Gamble Really Teaches

On September 5 a coin that did not exist six weeks earlier briefly traded at 1,799 dollars. Around the same time, mining machines built in 2018 for a storage network were earning close to 880 dollars a day pointing at it, and a day later the same machines earned closer to 380 as the crowd arrived. The coin is BTCB2, and if the name BIP 110 rings a bell, it should. This is the same fork we wrote about in August, the one that won 2.53 percent of Bitcoin's mining vote, mined two blocks, and stalled against difficulty math. It came back on August 30 with a new engine. What follows is what actually changed, what the numbers say now, and the one thing holders should refuse to do.

Where this chain came from

Quick recap for anyone joining late. BIP 110 proposed one year of strict limits on arbitrary data inside Bitcoin blocks. It needed 55 percent of hashrate to signal yes and got 2.53. A tiny chain enforced it anyway on August 8, inherited the full network's difficulty with a fraction of its power, and froze almost immediately. Our first article ended by noting that the people behind it had not changed their minds, and that code for a proof of work change existed with no date attached. The date turned out to be August 30.

On that day the stalled chain restarted from block height 961,640 with two changes and one carryover. The proof of work moved from SHA256d to BLAKE2b, permanently. The block size shrank to roughly 300 kB serialized. And the original data limits carried over, set to expire on September 1, 2027. The node software is a build of Bitcoin Knots, the same ecosystem that backed the original proposal. Changing the proof of work dissolved the difficulty trap that froze the first attempt, because the new algorithm meant a fresh difficulty schedule that adjusts to whatever hashpower actually shows up.

Why change the engine at all?

Start from first principles. A proof of work algorithm does not just secure a chain. It defines who gets to vote on it, because only machines that speak the algorithm can mine. BIP 110 lost its vote inside the SHA256d electorate, where the fleets that made money on data transactions hold the overwhelming majority. You cannot win that room. So the fork chose a different room. BLAKE2b has been the algorithm of the Sia storage network since 2015, and it comes with its own installed base of purpose built machines, the Antminer A3 and the Goldshell SC5 Pro among them. The fork's own documentation is candid about this. Those machines were supported from the first block.

Notice what that means. The fork did not escape mining hardware politics. It swapped a fleet it could never persuade for a smaller fleet it might. The entire BLAKE2b fleet on earth produces about 31 PH per second. Bitcoin routinely produces roughly thirty thousand times that. Shrinking the electorate is the whole move, and it worked in the narrow sense. The chain now mines blocks on schedule, something the August version never managed.

Three numbers decide whether this chain lives

First number. The world supply of BLAKE2b hashrate is about 31 PH per second, almost all of it built to mine Sia. That is the total electorate the fork moved into. Second number. BTCB2 itself reports around 5.1 PH per second after climbing 30 percent in a week, spread over roughly sixty pools and solo miners. In six weeks the fork has pulled about one sixth of the planet's BLAKE2b capacity. Third number. A Goldshell SC5 Pro pointed at the chain earned about 880 dollars on September 2 and closer to 380 a day later, as difficulty caught up with the gold rush. That decay curve is the market doing its job. Machines flow toward the higher paying chain until daily revenue converges, the same hashprice arithmetic that governs everything we sell.

Now put the first two numbers together, because this is the part a hashrate audience should see instantly. A majority attack on BTCB2 today needs about 5.2 PH per second. That is one sixth of the world's BLAKE2b fleet, and BLAKE2b rents openly on hashrate marketplaces. Against Bitcoin, no rental float on earth comes within shouting distance of a majority. Against a five PH chain, the float is the majority, many times over. This is the standing tradeoff of every minority chain that shares its algorithm with a larger market. The shared fleet that gave it instant miners is also a loaded weapon pointed at it around the clock. Its security budget is whatever it pays miners per day, measured against what the rental market charges for the same power. Put a price on that with the September 3 numbers. A machine doing about twenty terahash earned roughly 380 dollars a day on this chain, and no rational owner rents power out for much less than it earns. Scale that to 5.2 PH and majority control of BTCB2 prices out near a hundred thousand dollars a day, or about four thousand an hour. That is not a budget that needs a government. It is a card limit.

If you hold bitcoin, the biggest risk is not mining at all

Every holder of pre fork bitcoin technically holds the same coins on BTCB2. That sentence has moved money into scammers' pockets after every fork since 2017, and it is already happening here. Two things make this fork sharper than most. Replay protection exists but is optional. The chain added a new signature format that marks a transaction for one chain only, but ordinary signatures remain valid on both chains. Spend your bitcoin with a normal wallet and that same signed transaction can be rebroadcast on the fork, and in principle the reverse. And because no major exchange lists BTCB2 and no mainstream wallet supports it, a wave of claim tools has appeared to fill the gap. Their business model is your seed phrase.

• Never type your seed phrase into any BTCB2 wallet or claim tool. There is no legitimate claim flow that needs your bitcoin seed
• If the forked coins matter to you, move your bitcoin to fresh keys first with an ordinary transaction, and only then experiment from the emptied keys
• Treat any instant claim or instant swap offer as hostile until proven otherwise. Thin liquidity on one small venue is not an exit
• If you do nothing at all, you lose nothing on the main chain. Doing nothing is a fully respectable strategy

What this means if you buy hashrate

An honest boundary first. IonGrid routes SHA256d hashrate. BLAKE2b is not on our menu, and this story is a good part of why we move slowly on small shared algorithm chains. Delivering hashrate to them is easy. The hard part is that the rental market itself is their security model, and we would rather explain that plainly than sell into it quietly. What we do keep live is the machinery for reading stories like this one. The hashprice logic that explains the 880 to 380 decay is the same logic on our index page. The difficulty mechanics that froze the first fork and revived the second are the same ones we walk through in our difficulty explainer.

• The live hashprice index, the arithmetic behind the gold rush and the decay
• Network difficulty explained, the math that froze fork one and revived fork two
• The first BIP 110 article, the vote and the stall this chain grew out of
• Solo mining, where your machines vote with your own block template

So is it a coin or a statement?

The sober reading of six weeks. Sixty pools is real participation and one sixth of an algorithm's world supply is not nothing. It is also one small exchange, about a million dollars of trading on the wild day, no listings on the major price aggregators, and a price that moved 1,799 dollars mostly because almost nobody could sell. The first fork failed at arithmetic. This one fixed the arithmetic by construction, which means it now faces the harder test the original never reached. Not can it mine blocks, but does anyone need them. That test has no mandatory signaling and no deadline. It just runs.

So we end, as last time, with a question. When you cannot change the rules of the room, is moving to a smaller room a victory or an exile? BTCB2 is one of the only live experiments on that question in Bitcoin's history, and whatever your answer, it is worth watching honestly rather than cheering or jeering. Our hope is unchanged from August. That these fights keep playing out in the open, in signals and difficulty and prices anyone can verify, and that everyone whose machines cast a vote, on any chain, casts it knowingly.

Sources

Primary references, current as of our research date of September 10, 2026.

• The fork's own FAQ, source for the restart height, block size, and replay design
• Bitcoin.com News on the 1,799 dollar spike and the thin liquidity behind it
• Gate Research on BTCB2 hashrate, pool count, and the ecosystem gap
• BitInfoCharts, live Sia network hashrate, the world BLAKE2b supply
• MillionMiner on Sia machine earnings during the September gold rush
• Cryptolexicon's warning on seed phrase harvesting by fake BTCB2 wallets