DayOne, a Singapore-headquartered data centre operator valued at roughly $20 billion, has partnered with Melbourne-based Cortical Labs to build what both companies call Singapore’s first biological data centre. The facility will use living human neurons, not silicon, to process information.

The announcement marks the most commercially ambitious deployment yet for Cortical Labs, whose CL1 biological computing platform grows approximately 200,000 neurons from donated blood stem cells and interfaces them on silicon chips. Each unit reportedly consumes less power than a handheld calculator. If those efficiency claims hold at scale, the implications for AI infrastructure are significant: the industry’s single largest bottleneck is energy, and a computing substrate that operates on microwatts rather than kilowatts would fundamentally change the economics.

From Doom to Data Centres

Cortical Labs first attracted attention in 2022 when it demonstrated that a dish of cultured neurons could learn to play the video game Doom. That experiment was a proof of concept for a broader thesis: biological neurons, shaped by four billion years of evolution, are inherently more efficient information processors than transistors. The company has since packaged that insight into the CL1, a contained “body-in-the-box” system designed for rack-mounted deployment in commercial data centre environments.

The Melbourne prototype facility is already operational. Singapore is the first international expansion, and the choice of partner reflects the scale of ambition. DayOne raised over $2 billion across Series A and B rounds in early 2025 and is reportedly targeting a US IPO at a valuation of up to $20 billion. The company operates hyperscale campuses across Singapore, Southeast Asia, Japan, and Europe. It is not a research outfit experimenting with novel compute. It is an infrastructure operator making a commercial bet.

The Singapore Deployment

The initial phase will begin at the National University of Singapore’s Yong Loo Lin School of Medicine, where Cortical Labs will deploy a single rack of 20 Cortical Cloud units. NUS Medicine will collaborate on cultivating the biological neurons and validating the platform for research applications including neuro-inspired AI, biomedical modelling, and healthcare use cases.

From there, the partnership is structured to transition into a live deployment within a DayOne commercial data centre facility. The parties are exploring a phased expansion that could eventually reach 1,000 biological compute units in Singapore, subject to technical validation and regulatory approvals.

One thousand units, each containing roughly 200,000 neurons, would mean 200 million living human neurons processing workloads inside a commercial data centre. That is an extraordinary sentence to write in 2026.

Why Singapore

The timing aligns with Singapore’s evolving data centre policy. The government has been expanding capacity while tightening sustainability requirements, making over 200 megawatts available in its DC-CFA-2 zone under the new Green Data Centre Roadmap, which enforces higher energy efficiency standards. A computing platform that delivers meaningful throughput at a fraction of conventional power consumption would be a natural fit for a regulatory environment that is simultaneously hungry for compute and constrained on energy.

DayOne CEO Jamie Khoo framed it directly: “Singapore is raising the bar for sustainable data centre growth, and the market is responding with new approaches.” Cortical Labs CEO Hon Weng Chong positioned the partnership as offering “policymakers and industry a practical alternative: a sustainable pathway to AI adoption.”

What This Means for BCI

Cortical Labs sits at an unusual intersection. It is not a brain-computer interface company in the clinical sense — it does not implant anything into patients. But its core technology, the ability to grow, maintain, and computationally interface with living neurons at scale, is directly relevant to the BCI field.

The challenge of keeping neurons alive and functional outside the body, of reading and writing signals to biological tissue reliably, and of scaling those processes from lab bench to commercial deployment are exactly the challenges that clinical BCI companies face. If Cortical Labs can demonstrate that 200,000 neurons can perform useful computation inside a rack-mounted box for sustained periods, it validates a set of bioengineering capabilities that the entire neurotech industry depends on.

There is also a more speculative connection. As biological compute matures, the boundary between “computer that uses neurons” and “brain-computer interface” may become less distinct. A system that interfaces silicon with living neurons for the purpose of information processing is, in a meaningful sense, already a brain-computer interface. It is just not attached to a person.

The Open Questions

The efficiency claims are dramatic but unverified at scale. Running 20 neurons on a bench is different from running 1,000 units in a tropical data centre where temperature, humidity, and biological contamination are constant threats. The neurons need nutrients, waste removal, and precise environmental control. At data centre scale, the life support systems for biological compute may consume more energy than the neurons themselves save.

There is also no public benchmark comparing CL1 throughput to conventional hardware on equivalent workloads. “Less energy than a calculator” is a compelling headline, but the relevant metric is performance per watt on tasks that matter, and Cortical Labs has not yet published those numbers.

Regulatory questions remain open too. Singapore has no existing framework for facilities that house large populations of living human cells for commercial computing purposes. The NUS Medicine partnership suggests both parties understand that regulatory groundwork will need to be laid alongside technical validation.

None of these questions invalidate the project. But they distinguish announcement from achievement. The partnership between DayOne and Cortical Labs is significant because it represents the first time a major infrastructure operator has committed capital to biological computing. Whether the technology delivers on its promise will depend on what happens inside that first rack of 20 units at NUS.