Bitcoin broke into two chains on 8 August 2026. The trigger for the split was the soft fork proposal BIP-110. Its minority chain came to a standstill after two blocks, while the main chain ran on unchanged.
BIP-110 aims to cap the volume of non-financial data in Bitcoin transactions for one year. For example, that covers the data from Ordinals inscriptions. The soft fork limits data fields to 256 bytes and also caps OP_RETURN outputs at 83 bytes. To take effect without a split, it moreover requires a quorum of 55% of blocks within a two-week signalling window. The dispute behind it dates back to 2025. Back then, Bitcoin Core lifted the existing relay limit for OP_RETURN with version 30, which produced BIP-444 and later BIP-110. Strategy chairman Michael Saylor came out against the proposal in an essay in July 2026. In the end, only 2.53% of blocks in the last signalling period before the split voted in favour.
Subscribe to our newsletter
The best articles of the week, directly delivered into your mailbox.
AntPool block without signal splits Bitcoin into two chains
The two-week signalling window began at block 961,632. From that height, nodes running BIP-110 rules rejected every block that failed to set the signalling bit. That bit is a marker in the block header which miners use to indicate support for a proposal. The first block without the marker then came from the mining pool AntPool. The main chain accepted it. BIP-110 nodes, however, discarded it and followed an alternative block from the mining group Roughnecks. Two Bitcoin chains have run in parallel ever since. Their shared history therefore ends at block 961,631.
Roughnecks mines through the Ocean pool and appeared at the split as the "110th Mining Division". A second block followed, then production stopped. The chain has added nothing since block 961,633. Meanwhile, the main chain kept producing blocks without interruption. For its users, nothing changed. Anyone running a node with BIP-110 rules now sees a different chain from the rest of the network.
A regular activation would have required broad miner support. BIP-110 sets the threshold at 1,109 of 2,016 blocks, so 55% within the window from block 961,632 to 963,647. Yet only 51 blocks signalled in the period before the split, equal to 2.53%. A soft fork becomes binding in Bitcoin only when enough miners enforce it. Support was thus far below the requirement well before the split. The proponents pushed the separation through anyway.
Miner boycott freezes the BIP-110 minority chain
The minority chain's deficit grew by the hour. On the evening of 8 August it stood at seven blocks. By the morning of 9 August it had widened further, through 18 and 26 to a total of 57 blocks. Later the same day it reached 88, and at 15:27 UTC it hit 111 blocks. The main chain had reached block 961,744 by then, while the BIP-110 chain remained at 961,633. Measured against a ten-minute block interval, that equals a good 18 hours of lag. Until press time, no newer data point emerged.
Still, the standstill does not trace back to an outside attack. The miners withdrew on their own. In the first signalling period after the split, not one of 113 blocks signalled for BIP-110. Before that, the figure had at least reached 2.53%. At the same time, Ocean's hashrate shown on the BIP-110 chain collapsed from around 36 EH/s to roughly 1.25 EH/s. That is a 96.5% drop within a day. Roughnecks halted its mining operation on 9 August at 03:40 UTC. The group also called on other BIP-110 miners to pause.
"We don't see this as a defeat for the BIP-110 movement, but as an escalation to the next step." - Roughnecks, operator of the BIP-110 mining division
Established hashpower, by contrast, stayed united on the main chain. Foundry, F2Pool, AntPool, ViaBTC and MARA likewise kept mining there throughout. Nobody announced this boycott. It consisted purely of staying on the old chain. An attack on the new chain was therefore unnecessary, because it simply lacked the computing power.
Exchanges and custodians ignore the minority chain
Infrastructure ultimately decides the economic relevance of a forked chain. It needs working wallets, lasting miner support, exchange listings, a custody connection and buyers. So far, the BIP-110 chain lacks every one of these building blocks. A chain without such connections stays technically alive but economically irrelevant.
Neither Coinbase nor Kraken, Binance, BitGo, Fireblocks or Anchorage have publicly announced support. Coinbase and Kraken instead reported normal operations on the main chain in their status feeds. No major provider has listed a separate token for the minority chain either. A tradable counterpart to the main chain's BTC never emerged. Anyone holding Bitcoin through an exchange or a custodian consequently stayed on the main chain throughout. Institutional holders therefore faced no need to act. Michael Saylor put the share of hashpower remaining with the main chain at around 99.85%.
The market reacted just as soberly. Bitcoin traded at around USD 65,000 on the day of the split. The fork itself caused no visible move. In the following days, the price held roughly at that level. The last major soft fork ran comparatively smoothly. Taproot activated in November 2021 at block 709,632 with broad miner support and without a lasting minority chain. Back then, hashpower stood behind the proposal. This time it stands against it.
Data limits drive the dispute over the Bitcoin BIP-110 split
At its core, the conflict turns on which data belong in a Bitcoin block. BIP-110 proponents want to push back non-financial content such as Ordinals inscriptions for one year through a consensus rule. The starting point, however, was a change to the relay rules of Bitcoin Core, not a consensus rule. In the proponents' view, that loosening invited more spam onto the blockchain. The other side, in contrast, countered that it merely mirrored the real behaviour of miners. Saylor set out his position in a 110-point essay in July 2026. In his view, consensus rules should address proven security risks rather than the perceived purpose of a transaction.
The regular activation path lies out of reach for the minority chain. It would first have to reach block 963,648. That leaves it around 2,000 blocks short, and it has produced two so far. The new rules would apply from block 965,664 for 52,416 blocks, so roughly one year. At this pace, that point is unreachable.
So the proponents are working on a more radical option. Luke Dashjr maintains the Bitcoin implementation Knots. Together with the pseudonymous developer Dathon Ohm, he proposed switching the proof-of-work algorithm of the minority chain. Furthermore, they named RandomX, KT256, BLAKE3 and Scrypt as candidates. Such a switch would free the chain from existing SHA-256d ASICs. The silent boycott by the large pools would come to nothing. Still, the chain would have to build its own hardware base. BIP-110 remains a proposal without a chain to enforce it.
