
FutureBit demonstrated a web-based simulation using a fruit fly brain model to perform Bitcoin mining calculations, claiming potential energy efficiency 10 times greater than current 3nm ASICs if scaled to organic neurons, though the current experiment runs on conventional hardware and lacks competitive hash rate.
AI-generated summary
Bitcoin mining relies on specialized hardware called ASICs to perform SHA-256 calculations efficiently. As mining difficulty increases, miners seek more energy-efficient solutions, prompting experimentation with alternative computing approaches.
As Bitcoin miners turn to increasingly specialized chips for greater speed and energy efficiency, hardware maker FutureBit says it has tested a different approach: using a simulated fruit fly brain to perform mining calculations in a web browser.
The HashFly demonstration simulates activity across 2,914 neural pathways taken from MaleCNS v1.0, a digital wiring diagram of an adult male fruit flyâs brain and central nerve cord.
FutureBit said on X that the company is working to simulate all the neurons in the dataset with SHA-256 and plans to publish its findings.
âFun fact if this could be scaled on real organic neurons, it would hash at ~ 1 watt per terahash...10x the efficiency of the best silicon 3nm ASICs!â the company wrote.
This means that a mining system made with living neurons could perform one trillion calculations per second while consuming about one watt of electricityâroughly 10 times the efficiency of leading 3-nanometer ASIC miners. The estimate applies a fruit flyâs total power consumption to a hypothetical system in which every neuron mines Bitcoin continuously.
While inspired by biological neurons, the experiment runs on conventional computer hardware and lacks the hash rate and difficulty required to compete with Bitcoin miners.
Bitcoin miners compete to add transaction blocks by repeatedly running SHA-256 calculations, with hash rate measuring the number of attempts per second. HashFly lets users adjust the target but operates at a fraction of Bitcoinâs current mining difficulty; by comparison, FutureBit says its five-inch Apollo III ASIC miner reaches 18 terahashes per second.
A separate project called FlyMiner uses a digital map of 139,255 fruit fly neurons and 16.8 million connections to control a standard Bitcoin mining program. When simulated signals from the flyâs movement-control neurons reach a set threshold, the program tests possible solutions to a mining problem supplied by solo mining pool CKPool, at speeds of up to 700,000 attempts per second.
The fruit fly projects are the latest in a series of unconventional Bitcoin mining experiments.
AI outlook â possibilities, not facts
FutureBit will publish detailed findings on the HashFly project's simulation results
Likely · Within weeks

Blockchain security firm SlowMist discovered malicious code in versions 1.1 and 1.2 of the FomoPeek iPhone app, which was distributed via Apple's App Store and marketed as a cryptocurrency transaction tracker. The app contained hidden modules enabling iOS sandbox escapes to steal private keys and sensitive data, leading to approximately $579,900 in USDT theft traced to attacker address 0x6d37f2C5e8F8546b648D317295565dA95975f4BB. Crypto platforms including Binance and OKX have warned users to remove the app, update iOS, and move assets to new wallets.

The UN Security Council will hear briefings from AI executives including Sam Altman, Dario Amodei, Yoshua Bengio, and Clément Delangue on risks posed by artificial intelligence, such as autonomous cyberattacks, election interference, and weapon design. The session, organized by France, also includes invited statements from Chinese firms DeepSeek and Moonshot. Amodei advocated for slowing AI development and restricting chip exports to China, while Trump dismissed AI risks as a 'hoax'.

White-hat actors have moved 40.71 BTC worth $3.31 million from the Coldcard hardware wallet exploit into a recovery effort labeled 'Crypto Recovery Trust,' according to Galaxy Research. The funds represent 2.8% of the total $130 million theft stemming from a 2021 firmware flaw that generated weak seed phrases on Coinkite devices.

OpenAI released GPT-6 Sol and GPT-6 Luna models on Tuesday, minutes after Anthropic launched Claude Opus 5.5. Sol and Luna are positioned as cheaper, faster alternatives to GPT-6 Astra, with API costs reduced by 50% compared to GPT-5.6 promotional rates. OpenAI claims Sol outperforms Claude Opus 5 on AutomationBench and Agents' Last Exam at significantly lower cost per task, and both models are now available in ChatGPT Work and Codex for paid tiers, with Luna also reaching free users via desktop app.

Anthropic released Claude Opus 5.5, stating it performs at the level of its priciest flagship model Fable 5.1 on most tasks while costing 20% less to run than Opus 5 and 60% less than Fable 5.1. The model leads in agentic coding benchmarks like Terminal-Bench 4.0 and FrontierCode, and is now live across AWS, Google Cloud, and Azure, with Sonnet 5.5 and Haiku 5.5 to follow.

White hats transferred 52.37 Bitcoin rescued from wallets exposed by a Coldcard vulnerability to the Wyoming-based Crypto Recovery Trust to return funds to victims.