BitcoinIII

The security of a proof-of-work (PoW) network depends on an attacker not being able to independently amass more hashrate than the network as a whole. In the case of ASIC-mineable coins — such as Bitcoin — a coin can only be secure if its network is large enough and, crucially, if the vast majority of its dedicated mining hardware is being used to mine the coin itself.

Bitcoin, for example, is the first cryptocurrency to have ever been launched, and went live in January 2009. At the start, Bitcoin was CPU-mineable, later moving to GPUs, FPGAs, and, finally, ASICs. The ASICs that were manufactured were made specifically for Bitcoin itself. I.e., Bitcoin mining hardware developed into its own hardware class, completely separate from other hardware classes. Fast-forward to today, and this is still true: there are hundreds of purpose-built ASIC types for a plethora of different proof-of-work algorithms, but only those that run little-endian SHA-256d over Bitcoin's 80-byte block header are able to mine Bitcoin. Therefore, SHA-256d ASICs were developed for Bitcoin, and grew with Bitcoin. As a result, Bitcoin controls over 99% of global SHA-256d hashrate and is, therefore, one of the most secure networks on the planet. This is one of the main reasons Bitcoin is so valued — when someone broadcasts a transaction to the Bitcoin network, and it is included in a block, after a few confirmations, one can be certain that undoing the transaction is an impossibility.

The issue that arises is when smaller coins adopt SHA-256d as their PoW algorithm — or any PoW algorithm shared with a larger coin (though SHA-256d is the most fitting example). A small SHA-256d coin will be destined to compete with Bitcoin for the same hardware in perpetuity, unless it miraculously managed to gain the same value as Bitcoin and, consequently, "out-hashed" Bitcoin. Therefore, new SHA-256d coins are unlikely to ever gain a high-enough level of network security to attract meaningful capital, since users and investors will not invest in a coin that can have its history rewritten or block production controlled by hardware being used to primarily mine another coin.

BC3 recognised the above issues and changed the only aspect of consensus required to address them: the PoW algorithm. Block hashes are computed using SHA3-256t: 3 rounds of SHA3-256 over the 80-byte Bitcoin block header (a drop-in replacement for SHA-256d), and the PoW check is performed using the little-endian representation of the output hash. BC3 is the first coin to ever use this exact PoW algorithm.

By shifting away from SHA-256d (and any other currently ASIC-mineable PoW algorithm), BC3 ensures that it has the potential to develop into its own hardware class, rather than remaining in Bitcoin's shadow. If it ever becomes valuable enough for a fleet of ASICs to be built for it, that fleet would be built for BC3, the largest coin on the algorithm, and there wouldn't be any other "pool" of ASICs being able to be redirected towards the chain at short notice.

BC3 is a fork of BitcoinII (BC2) Core v29.1, itself a fork of Bitcoin Core v29.1. Apart from network parameters, the code is Bitcoin’s with a single consensus change: the block hash and proof-of-work algorithm is SHA3-256t — three iterations of SHA3-256 — instead of SHA-256d.

The algorithm is selected by version bit 12. Pre-fork blocks do not have bit 12 set; post-fork blocks must. This makes the hash algorithm an intrinsic property of the block header, leaves the BIP320 version-rolling bits and the top three BIP9 bits untouched, and keeps the lower twelve bits available for BIP9 signalling.

The hard fork activated at block height 30,240. Block 30,239 was the last SHA-256d block; block 30,240 the first SHA3-256t block. Difficulty was reset to 1 (nBits = 0x1d00ffff) at the fork height so the new network could find its hashrate without producing blocks 35 years apart.