Carbon-based computing challenges silicon-based GPUs? A complete fruit fly brain computational model drives a physical body to perform natural behaviors

According to CoinWorld, monitored by 1M AI News, the full-brain simulation company Eon Systems recently demonstrated the first closed-loop system driven by a complete brain simulation that controls a physical body: a full-brain model built from the connectome of real fruit flies driving a virtual fruit fly in the MuJoCo physics engine to perform natural behaviors such as walking and grooming. Previously, no one had achieved a biological connectome-derived full-brain simulation that could generate multiple natural behaviors to drive a physical body. The core of the system is the fruit fly whole-brain computational model published by Eon senior scientist Philip Shiu in Nature in October 2024. Shiu’s team used electron microscopy data from the FlyWire connectome and machine learning-predicted neurotransmitter types to construct a leaky integrate-and-fire model encompassing over 125,000 neurons and 50 million synaptic connections, relying solely on connectome topology, synaptic weights, and excitation/inhibition mapping, without neuron-by-neuron parameter tuning. The model’s predictions matched 91% of 164 published experimental observations. When synaptic weights were randomly shuffled, only 1 out of 100 simulations could reproduce the correct behavior, indicating a strong structural constraint of the connectome on brain function. This demonstration integrated the brain model into the NeuroMechFly v2 simulation framework, forming a complete closed loop from sensory input to whole-brain neural activity to motor output. Eon Systems is headquartered in San Francisco, with an advisory board including Harvard geneticist George Church and Mathematica founder Stephen Wolfram. The next goal is whole-brain simulation of mice (about 70 million neurons, 560 times that of a fruit fly), with the ultimate aim of simulating the human brain.