Singapore's First Biological Data Center Launches
Thailand Business News
Singapore just switched on its first data center that runs on lab-grown human neurons instead of just silicon chips. It's a real test of whether "biological computing" can cut AI's power appetite, not just a lab experiment anymore.
Based on reporting by Thailand Business News — read the original for the full story.
Summary, retelling and take written by AI under human oversight; images are AI-generated illustrations. How we work · Report an error
DayOne Data Centers, Cortical Labs and NUS Medicine's Yong Loo Lin School just flipped the switch on something genuinely strange: a data center prototype in Singapore that pairs conventional server infrastructure with living neurons grown from stem cells. The centerpiece is a 20-unit CL1 biological computing system, which the partners describe as the first independently operated biologically integrated server rack anywhere in the world. It's now sitting inside a working research environment at NUS Medicine, built and supported by DayOne's infrastructure team.
The pitch here isn't science fiction for its own sake. Those neurons take electrical signals, respond to them, and adapt over time — and they do it, according to the companies, on a fraction of the wattage that digital silicon needs for comparable tasks. That matters a lot in Singapore right now, where the government is pushing a Green Data Center Roadmap that demands higher energy efficiency even as AI and cloud demand keeps climbing. DayOne CEO Jamie Khoo framed it plainly: the goal is proving that scaling compute and cutting resource intensity aren't opposing forces.
Cortical Labs founder Hon Weng Chong is thinking past the lab bench already. He says biological computing's real edge shows up where data is scarce and conditions shift constantly — drug discovery, humanoid robotics, cybersecurity, fraud detection. Rather than replacing silicon outright, the idea is a complement that learns from less and adapts faster in those specific niches. Professor Rickie Patani of NUS Medicine, meanwhile, sees a second payoff beyond efficiency: a platform for studying how learning and adaptation actually work at the biological level, which could speed up neurological disease research and drug discovery in ways today's silicon systems can't.
This is one piece of a much bigger buildout for DayOne, which operates across nine markets in Asia Pacific and Europe and has booked roughly 2.1 GW of capacity since it was established in 2022. Its first Singapore data center, SG1, hit structural topping out in May 2026 and is due to go live in early 2027, alongside a Global Operations Command Center the company is also standing up in the country. The biological prototype is small in scale next to those numbers, but it signals where DayOne wants to plant its flag: not just building bigger data centers, but experimenting with fundamentally different ways of computing under the same roof.
My take — AI-written commentary, not fact-checked reporting
Neurons in a server rack sound like a gimmick until you remember the actual problem AI infrastructure has right now: power. Everyone's chasing efficiency gains from better chips and cooling, so a wetware approach that claims dramatically lower wattage for narrow, adaptive tasks deserves attention rather than eye-rolls. That said, this is a 20-unit prototype in one lab, not a blueprint for replacing silicon data centers, and the companies themselves are pointing to niche use cases like drug discovery rather than general AI workloads. Singapore backing this kind of experiment fits its pattern of using tight environmental rules to attract infrastructure bets other markets won't touch first.”
Read more about this at: Thailand Business News