Glacian sells $1.91m of a $5m round on megawatts recovered from cooling

Something at the site has to accept the setpoint. Glacian's software builds physics-based digital twins of cooling equipment and uses machine-learning models to recommend operating settings from conditions inside and outside the facility ^[8]. A recommendation becomes a saving only when an operator, or a building management system wired to take the value, acts on it. Integration through existing monitoring and data-center infrastructure management systems, across cooling hardware vendors ^[9], is where a retrofit attaches. The cooling plant is already bought.

The commercial claim is substitution. Power no longer required for cooling can support additional servers inside an existing power allocation ^[10]. For that to turn into compute, the recovered watts have to show up at peak outdoor conditions, when the allocation binds. A twin calibrated during mild weather will produce a percentage that does not hold in August.

The money is smaller than the headline number twice over. $1.91 million against a $5 million offering is 38.2 percent of the round ^[1][1]. Glacian said the total includes $150,000 of simple agreements for future equity that converted, money already disclosed in an August 4 filing ^[4], which leaves at most $1.76 million of new cash ^[2].

The founders have been working on this longer than the company has existed. Zhao and Zuo formed Glacian in January, about 13 years after meeting at a conference where Zhao was pursuing a doctorate at Carnegie Mellon and Zuo was teaching at the University of Miami, Zhao told Technical.ly ^[5]. Zhao then spent a decade at Delos, a New York property-technology business, eventually leading commercialization work there ^[6]. "We think this is the right timing, so I quit my job," Zhao told Technical.ly about leaving Delos in late 2025 ^[7]. Zuo is a Penn State professor of architectural engineering and a fellow of ASHRAE and the International Building Performance Simulation Association ^[12], and Glacian says the underlying work drew more than $8 million in research support beginning in 2013 from bodies including the Department of Energy, the Department of Defense, the National Science Foundation, ASHRAE, JPMorgan Chase and Penn State ^[13].

The demand case comes from a Department of Energy update estimating data centers could account for 11.8 percent of US electricity consumption by 2030, with scenarios ranging from 9.5 to 15.3 percent ^[14]. A national consumption share still leaves an operator asking whether the substation feeding one building has headroom this year. The spread across those scenarios is 5.8 percentage points ^[3], wider than most capacity plans tolerate.

Zuo announced an investment from Samsung Next, alongside a collaboration with Samsung Research America, on LinkedIn on July 22, without a check size or any indication in the Form D of how that investment maps to the disclosed proceeds ^[15]. The filing names no buyers and sets no valuation ^[11]. The company calls the software "Physical AI" ^[8].