Singapore's newest data center is a biological marvel, a living, breathing alternative to traditional silicon chip servers. This cutting-edge facility, housed within the National University of Singapore's Centre for Life Sciences, processes data using human brain cells, marking a significant shift in the field of data processing and artificial intelligence. The center, a collaboration between Cortical Labs, NUS, and DayOne, is a testament to the innovative spirit of Singapore's tech industry.
What makes this data center truly unique is its reliance on living neurons. Every three days, lab technicians provide the brain cells with a nourishing cocktail of sugar, micronutrients, and pH buffers, ensuring their vitality. A sophisticated gas mixer delivers a steady supply of carbon dioxide, oxygen, and nitrogen, creating a dedicated life support system. This biological approach is a stark contrast to the traditional silicon chip servers, which are energy-intensive and require extensive cooling.
Cortical Labs' founder and CEO, Chong Hon Weng, believes that biological data centers are ideal for applications with limited datasets and highly unpredictable conditions. He envisions humanoid robots navigating public spaces, homes, and offices, where traditional servers fall short due to the vast amounts of training data required. For instance, teaching a robot to navigate from one point to another in a real-world environment demands extensive training data, which is not always readily available.
The biological data center's capabilities extend to cybersecurity as well. Chong suggests that these centers can spot anomalies without relying on massive training datasets, mirroring how humans learn from a few examples and adapt to changing circumstances. This efficiency in learning and adaptability is a significant advantage over traditional servers.
Singapore's first biological data center houses 20 units of biological computers (CL1s), each containing at least 200,000 lab-grown neurons on an electrode-fitted silicon chip. These neurons, derived from blood cells reprogrammed into stem cells, communicate with a computer, translating their neural activity into raw computing power. The plan is to expand this facility to accommodate up to 1,000 CL1s, pending regulatory approval and energy efficiency and safety tests.
The Melbourne facility, currently operating 120 CL1s, boasts around 20 paying customers, including corporate research arms and universities experimenting with robotics and gaming. Customers pay a monthly fee of US$2,200 (S$2,800) to access the computing resources of a CL1, which is significantly lower than the US$4,300 monthly fee charged by major cloud platforms for renting high-end AI chips. However, Chong acknowledges that traditional silicon chips excel in fast, precise, and repeatable calculations, making them the preferred choice for large language models like ChatGPT.
One of the most compelling advantages of biological data centers is their minimal energy footprint. Unlike traditional servers, which generate heat and require cooling, brain cells use very little energy. Each CL1 unit consumes only 30 watts, which is less power than a handheld calculator, even with its life-support systems. In contrast, an AI silicon chip like Nvidia's H100 SXM can consume up to 700 watts during demanding tasks, and a typical data center server with eight H100 chips can consume a staggering 10,200 watts, including supporting hardware.
Singapore's energy and water constraints have led to a temporary pause on new data center construction since 2019, as the country aims for greener digital growth. Data centers accounted for 5.3% of the country's electricity consumption in 2019, rising to 7% in 2020 due to the pandemic's acceleration of digitalisation. Biological data centers, with their low energy demands, offer a sustainable solution to these challenges.
Chong highlights Singapore's role as a major fibre connectivity hub in the Asia-Pacific region, making it an ideal location for data centers. However, the country's energy and water constraints are significant. Cortical Labs' technology, he believes, can help overcome these challenges, and the NUS facility will play a crucial role in determining the manpower requirements and skill sets needed for commercial applications.
In conclusion, Singapore's biological data center is a groundbreaking innovation, offering a sustainable and adaptable alternative to traditional silicon chip servers. As the world grapples with the energy demands of data centers, this facility presents a compelling solution, raising questions about the future of data processing and artificial intelligence. It is a testament to the power of innovation and the potential for a greener, more efficient digital future.
