From 1.5GW bookings to a $20 billion valuation ambition, Singapore's data center operator DayOne is sprinting towards the US stock market.

According to reports citing informed sources, DayOne Data Centers Ltd., one of the key participants in the global data center construction boom, has secretly submitted its initial public offering (IPO) application to the U.S. stock market. The sources indicate that the Singapore-based data center operator plans to go public as early as next quarter. As the information has not yet been disclosed, the sources requested anonymity. They stated that the company has been considering raising approximately $5 billion through this IPO. Previously, reports indicated that DayOne might seek a valuation of about $20 billion through the listing. The sources noted that considerations are still ongoing, and detailsincluding fundraising size and listing timingmay change. DayOne did not immediately respond to requests for comment. In June, DayOne completed a $4.5 billion Series C funding round, led by its largest shareholders, Coatue Management and Hillhouse, along with new investors Achi Capital Partners and Indonesias sovereign wealth fund. The company stated that this funding will accelerate DayOne's large-scale expansion in key markets such as Singapore, Malaysia, Indonesia, Thailand, Japan, Hong Kong, Finland, and Spain. Since its founding in 2022, DayOne has secured over 1.5 gigawatts (GW) of data center capacity bookings in the Asia-Pacific and European regions. Its investors also include prominent data center operator GDS from China, as well as SoftBank's Vision Fund led by Masayoshi Son and Citadel founder Ken Griffin. With the rise of artificial intelligence technologies, investor interest in data centerscrucial infrastructure for supporting AI training and inferencehas surged. Australian company Firmus Technologies Pty announced last week that it has received investment commitments for a $2 billion funding round, with Coatue also participating. Other well-known large data center operators preparing for their U.S. IPOs include Switch Inc. and Nscale, which may go public as early as this year. Dallas-based Csquare Inc. raised $1.05 billion through an IPO last month, setting its offering price at $21 per share, lower than previously projected. This data center company, supported by North American asset management giant Brookfield Corp., closed at $21.60 on the New York Stock Exchange on Monday. The strong demand for AI computing power and the booming construction of data centers have created a flourishing landscape for AI investments, which is the core strategic reasoning behind DayOne's decision to sprint to the U.S. capital markets at this moment. DayOne is not relying on long-term story financing: the company just completed a $4.5 billion Series C equity financing in June, and since its founding in 2022, it has secured over 1.5 GW of capacity bookings in the Asia-Pacific and Europe, clearly aiming to build "AI-ready" data center parks for hyperscale cloud providers and enterprise clients. Therefore, this IPO is essentially about repositioning power + land + data centers + AI computing capacity, which has already been validated by private capital, in the largest public equity capital markets during the most robust capital expenditure cycle of the AI boom. What investors should focus on regarding DayOne's U.S. IPO is not simply "another AI concept stock going public," but rather that the AI bull market is causing an asset class diffusion: from Nvidia's GPU-dominated AI semiconductor profits, extending to data center capacity leasing cash flows, and further to the infrastructure equity, credit, and core securitization tools that support these assets. Switch has also recently secretly submitted its U.S. IPO application, followed closely by DayOne, indicating that Wall Street asset management giants are attempting to repackage AI data centers from high capital expenditure projects into digital infrastructure cash flow assets characterized by long-term contracts, capacity bookings, and potential residual value. From the perspective of large AI systems engineering, the enhancements in model performance have shifted bottlenecks from a single GPU to the entire physical infrastructure chain, encompassing "GPU clusters high-speed networks power access liquid cooling/thermal management data center parks long-term financing in financial markets." DayOne happens to be situated at the layer of "data centers + power capacity," which has the highest capital intensity, longest construction cycles, and strongest supply constraintsselling not the AI models themselves, but the available megawatts/gigawatts (MW/GW) of AI computing infrastructure capacity that can sustain tens of thousands, or even hundreds of thousands, of GPUs continuously running.