AI Card Bao HERMES.AICARDBAO.COM

Industry News

Elon Musk said that by 2029, the computing power available in orbit would far surpass that on Earth; the head of Google’s energy division responded by saying it would be 100 times greater.

There was less than two days between these two statements. On October 31, Musk said on X that the monthly deployment volume of orbital computing systems might exceed that of ground-based systems by around the end of 2028, and would significantly surpass them in 2029. “Compared to orbital computing, the processing power of ground-based systems is practically insignificant,” he added.

Amanda Peterson Corio, the person in charge of energy management at Google’s data centers, said at the WTF conference that the cost of deploying hardware in space can be as much as 100 times higher than that of building similar systems on Earth. Moreover, issues related to maintenance and upgrades in orbit must also be addressed. On Earth, it is possible to replace servers or rewire systems at any time, but such options are not available in space. “Before embarking on space missions, it is essential to explore all feasible solutions on Earth first.”

Musk’s post was in response to a forecast made by Sequoia Capital partner Shaun Maguire: SpaceX’s revenue would increase from $51.5 billion in 2026 to $731 billion in 2028. Maguire himself noted that there were significant uncertainties involved, and Sequoia Capital held a substantial amount of SpaceX shares. Musk argued that this forecast was actually rather conservative. The concept of “orbital computing power” is also one of the key factors underlying SpaceX’s market value of approximately $1.8 trillion; therefore, it is essential to consider the perspective of the person making such claims when interpreting their statements.

Google has not completely closed the door on this possibility. Its Suncatcher project has already sent prototypes equipped with custom-made TPUs into space. The paper suggests that the cost per kilogram should be reduced to below $200 in order to make such missions economically viable, and this goal is expected to be achieved around 2035, depending on the progress of the Starship project. There is a six-year gap between 2029 and 2035.

SpaceX has proposed deploying up to 1 million satellites in its constellation project, while Blue Origin’s plan calls for the deployment of a maximum of 51,600 satellites. Both proposals are still at the conceptual stage.

Peterson Corio also holds a different view: In fact, the United States has sufficient electricity supply; what really hinders development are the lengthy approval processes and extensive evaluation procedures required for connecting new power sources to the grid. If she is correct, then one of the main selling points of this space-based solution would be diminished in significance.

Related