March 10, 2026 Noah Conway Technology 0

Un petit nombre de start-up cherchent désormais à réaliser des calculateurs biologiques.

Les data Center classiqes consomment beaucoup d’énergie et de ressoures rares. Un petit nombre de start up cherchent aujourd’hui les remplacer par des data Center bioloique uilisant des neurones mumains. C’est le cas de la socié australienne Cortical Lab.

Mais ces centres sont difiles à construire et à maintenir comme l’explique Michael Barros University of Essex,Great Britain. Cortical Lab est le premier à y réussir.

A suivre, non traduit”

Although these systems can be trained for relatively simple tasks such as gaming Fatethe exact way these neurons work and how best to train them to perform tasks such as machine learning is still unclear, he says Reinhold Schereralso at the University of Essex. “Accessing this allows you to explore learning, training and programming,” he says. “You don’t program neurons like standard computers.”

Cortical Labs says its data centers will also require much less power than typical computing systems, saying each CL1 needs around 30 watts, rather than the thousands of watts required by a state-of-the-art conventional AI chip.

“If we scale it up and have them as whole rooms, like you have now with data servers, then there could be huge energy savings,” he says Paul Roach at Loughborough University in Great Britain. There are other resources that biological data centers may need, such as nutrients to nourish and keep neural chips alive, but it should require much less cooling than conventional computing, he says. “The amount of energy saved [Cortical Labs’s] the numbers are quite conservative.”

However, the technology is still in its early stages, he said Tjeerd old Scheper at Oxford Brookes University in the UK, who collaborated with a rival biocomputing company, FinalSpark. “Will it work as well as people might think? No, we’re still in the early days of this development.”

It’s hard to make a direct size comparison because the CL1 chips can’t perform conventional calculations like a regular silicon-based AI chip, but the proposed biological data center will have hundreds of biological chips compared to the hundreds of thousands of graphics processing units (GPUs) seen in the largest AI data centers.

“I think it’s a very long way from being production ready. It’s a very big step from a small network playing computer games to an LLM,” he says Steve Furber at the University of Manchester in the UK.

One of the remaining problems is that it is still not clear how to store the results of training these neurons in the form of memory, or how to run actual computational algorithms on them, rather than training them for specific uses such as video games.

Another challenge is how to retrain neurons once they have completed a specific task. “Everything they’re trained in is lost when the culture ends its life, so there needs to be proper retraining,” says Scherer. “Then it’s not an optimal solution to keep the technology going if you need to retrain every 30 days.”

topics: