Scientists at the Jülich Research Centre in Germany are preparing to simulate the human brain at full scale using one of the world’s most powerful supercomputers, marking a major leap in computational neuroscience.
Building on earlier breakthroughs — including the 2024 mapping of a fruit fly’s brain circuitry containing 54.5 million synapses — the researchers aim to combine multiple models of smaller brain regions into a single large-scale simulation capable of handling billions of firing neurons.
Led by neurophysics professor Markus Diesmann, the team plans to run the model on the JUPITER supercomputer, currently ranked fourth globally, which features thousands of graphical processing units and exascale computing capabilities.
The researchers recently demonstrated that a “spiking neural network” could be scaled to match the human cerebral cortex’s estimated 20 billion neurons and 100 trillion connections, a development they say sets this effort apart from earlier projects such as the Human Brain Project. “We know now that large networks can do qualitatively different things than small ones,” Diesmann told New Scientist. “It’s clear the large networks are different.”
However, experts caution against overstating the implications, noting that even a full-scale simulation has limits. “We can’t actually build brains,” University of Sussex mathematical physics professor Thomas Nowotny said. “Even if we can make simulations of the size of a brain, we can’t make simulations of the brain.”