A global team harnessed the Fugaku supercomputer to build one of the largest and most realistic virtual mouse cortex models ever created.
The digital brain behaves like a real one, allowing scientists to study disease progression, neural dynamics, and potential treatments safely in silico. Combining biological datasets with powerful modeling software, the international team proved that large-scale, biophysically accurate brain models are now feasible.
Supercomputer-Powered Whole Cortex Simulation Breakthrough
Harnessing the power of one of the fastest supercomputers on Earth, researchers have created one of the most expansive and detailed biophysically realistic brain simulations of an animal to date. This fully digital version of a mouse cortex gives scientists a new way to explore how the brain works by recreating conditions such as Alzheimer’s or epilepsy within a virtual environment. They can observe how damage moves through neural networks or examine processes related to cognition and consciousness. The model captures both structure and activity, incorporating nearly ten million neurons, 26 billion synapses, and 86 connected brain regions.
This achievement was made possible by Supercomputer Fugaku, Japan’s flagship system capable of performing quadrillions of calculations per second. Teams from the Allen Institute, along with Tadashi Yamazaki, Ph.D., of Japan’s University of Electro-Communications and three additional Japanese institutions, collaborated to bring the project to life. The full details will be presented in an upcoming paper at SC25, the leading global supercomputing conference that will take place in mid-November.
Fugaku and Global Team Drive a New Neuroscience Frontier
Scientists can now use this model of the mouse cortex to explore how diseases develop, how brain waves support attention, or how seizures move across the brain. Previously, studies like these required real biological tissue and could only be conducted one experiment at a time. With this new approach, researchers can investigate many ideas in a virtual space. The simulations may help uncover early signs of neurological disorders before symptoms appear and enable scientists to safely evaluate potential treatments or therapies in a digital setting.
“This shows the door is open. We can run these kinds of brain simulations effectively with enough computing power,” said Anton Arkhipov, Ph.D., an investigator at the Allen Institute who worked on the project. “It’s a technical milestone giving us confidence that much larger models are not only possible, but achievable with precision and scale.”
This international effort brings together deep neuroscience expertise and advanced computing capabilities. The Allen Institute provided the framework and biological details for the digital brain using data from the Allen Cell Types Database and the Allen Connectivity Atlas, and Japan’s Fugaku executed the computations that turned that information into a functioning simulation.
Merging Biological Data With Massive Compute Power
Fugaku, jointly developed by RIKEN and Fujitsu, is one of the world’s fastest supercomputers capable of more than 400 quadrillion operations per second. To put that into perspective, if you started counting right now, one by one per second, it would take over 12.7 billion years to reach that number (approximately the age of the universe: 13.8 billion years). “Fugaku” comes from Mount Fuji, and just like the mountain’s high peak and broad base, it was chosen to symbolize its power and wide reach.
“Fugaku is used for research in a wide range of computational science fields, such as astronomy, meteorology, and drug discovery, contributing to the resolution of many societal problems,” said Yamazaki. “On this occasion, we utilized Fugaku for a neural circuit simulation.”
The supercomputer is made up of small parts called nodes, which are grouped together in layers like units, shelves, and racks. Together, these components add up to a total of 158,976 nodes, allowing Fugaku to manage a massive volume of data and computations.
Using the Allen Institute’s Brain Modeling ToolKit, the team translated data into the working digital simulation of the cortex. A neuron simulator, Neulite, turned equations into neurons that spike, signal, and chatter just like their living counterparts.
Turning Real Data Into a Biophysically Accurate Virtual Cortex
Watching a simulated mouse cortex is like watching biology in real time. It captures the actual structure and behavior of brain cells, down to the branches coming from neurons, the activations of synapses—the tiny contacts conveying messages from upstream neurons to the branches of downstream neurons—and the ebb and flow of electrical signals across membranes. “It’s a technical feat, but it’s only the first step,” said Yamazaki. “God is in the details, so in the biophysically detailed models, I believe,”
“Our long-term goal is to build whole-brain models, eventually even human models, using all the biological details our Institute is uncovering,” said Arkhipov. “We’re now moving from modeling single brain areas to simulating the entire brain of the mouse.” With this kind of computational power, the goal of a full, biophysically accurate brain model isn’t just science fiction anymore. Scientists are in a new frontier where understanding the brain means, quite literally, being able to build one.
Meeting: SC25
This cutting-edge research was made possible by an international team including Laura Green, Ph.D.; Beatriz Herrera, Ph.D.; Kael Dai, B.Sc.; Rin Kuriyama, M.Sc.; and Kaaya Akira, Ph.D.
Never miss a breakthrough: Join the SciTechDaily newsletter.
Follow us on Google and Google News.
1 Comment
It’s becoming clear that with all the brain and consciousness theories out there, the proof will be in the pudding. By this I mean, can any particular theory be used to create a human adult level conscious machine. My bet is on the late Gerald Edelman’s Extended Theory of Neuronal Group Selection. The lead group in robotics based on this theory is the Neurorobotics Lab at UC at Irvine. Dr. Edelman distinguished between primary consciousness, which came first in evolution, and that humans share with other conscious animals, and higher order consciousness, which came to only humans with the acquisition of language. A machine with only primary consciousness will probably have to come first.
What I find special about the TNGS is the Darwin series of automata created at the Neurosciences Institute by Dr. Edelman and his colleagues in the 1990’s and 2000’s. These machines perform in the real world, not in a restricted simulated world, and display convincing physical behavior indicative of higher psychological functions necessary for consciousness, such as perceptual categorization, memory, and learning. They are based on realistic models of the parts of the biological brain that the theory claims subserve these functions. The extended TNGS allows for the emergence of consciousness based only on further evolutionary development of the brain areas responsible for these functions, in a parsimonious way. No other research I’ve encountered is anywhere near as convincing.
I post because on almost every video and article about the brain and consciousness that I encounter, the attitude seems to be that we still know next to nothing about how the brain and consciousness work; that there’s lots of data but no unifying theory. I believe the extended TNGS is that theory. My motivation is to keep that theory in front of the public. And obviously, I consider it the route to a truly conscious machine, primary and higher-order.
My advice to people who want to create a conscious machine is to seriously ground themselves in the extended TNGS and the Darwin automata first, and proceed from there, by applying to Jeff Krichmar’s lab at UC Irvine, possibly. Dr. Edelman’s roadmap to a conscious machine is at https://arxiv.org/abs/2105.10461, and here is a video of Jeff Krichmar talking about some of the Darwin automata, https://www.youtube.com/watch?v=J7Uh9phc1Ow