FutureBit introduced HashFly on September 13, 2026: a browser-based proof of concept that runs a simulated fruit-fly connectome against a simplified Bitcoin-mining target. The activity models part of the mining process; it is not mining on Bitcoin’s live network.
What HashFly simulates
HashFly identifies its dataset as MaleCNS v1.0 and lists 2,914 reconstructed traces. The broader connectome is reported to contain 165,122 reconstructed neurons, and the demonstration does not use the full 25-million-synapse graph.
The site describes a sequence in which simulated R1–R6 photoreceptors read a block header. SHA-256d—a double application of the SHA-256 hashing function—is checked against a toy target, meaning a simplified threshold for the demonstration. If the hash falls below that threshold, two PPL101 cells respond in the simulation.
How the simulated mining sequence works
HashFly’s description connects the block header, the hash calculation, and the simulated cellular response in three steps.
What the browser run recorded
In one reported browser run, about 80 blocks were checked in roughly an hour, with a hash rate of about 100 kH/s. Those figures describe activity in the demonstration, whose maximum difficulty is level 6—far easier than finding a real Bitcoin hash. They are not counts of Bitcoin-network blocks or evidence of a payout.
FutureBit’s organic-neuron estimate is conditional
FutureBit said a system scaled to real organic neurons could reach about 1 W/TH and be ten times as efficient as the best silicon 3 nm ASICs. That projection describes a hypothetical system using real neurons; HashFly itself is a simulation, and the projected efficiency is not a measured result from the browser demo.
