Biocomputing Moves From Lab to Data Center
A startup is constructing what it claims will be the world's first data center powered by human brain cells, marking the most ambitious attempt yet to turn biological computing from a research curiosity into practical technology. The company plans to use organoids, clusters of lab-grown human neurons, as core processing units in a facility designed to handle real computational workloads.
The approach exploits a fundamental advantage that biological neural networks hold over silicon chips: energy efficiency. The human brain performs roughly 10 quintillion operations per second while consuming only about 20 watts of power, less than a typical light bulb. Modern data centers consume megawatts of electricity and require elaborate cooling systems that add further energy costs.
How Brain-Cell Computing Works
The technology builds on a decade of advances in organoid research. Scientists have learned to grow clusters of human neurons in laboratory dishes, where the cells self-organize into three-dimensional structures exhibiting electrical activity resembling brain function. These organoids form synaptic connections, process signals, and display rudimentary learning behavior.
In a biocomputing context, organoids are interfaced with electronic systems through arrays of microelectrodes that both stimulate neurons and read their electrical responses. Input data is encoded as patterns of electrical stimulation, the organoid processes these signals through its neural network, and the output is read back through the electrode array.
Previous demonstrations have shown organoids can learn to play simple video games, recognize patterns, and perform basic classification tasks. The startup aims to scale this by deploying thousands of organoids in parallel, each handling a portion of a workload, with conventional electronics managing coordination and data routing between biological processing units.
