Musk predicts: orbital computing power will far exceed ground computing power by 2029

Article body EM_StockImg_Start

EM_StockImg_End

SpaceX CEO and Tesla CEO, the world's richest person Musk, latest prediction: by 2029, the artificial intelligence (AI) and cloud computing processing capacity deployed in low Earth orbit will far exceed ground-based computing power.

Musk posted on social media platform X on October 31 that the monthly deployment volume of orbital computing may exceed ground-based computing power around the end of 2028, and will substantially exceed ground-based computing power in 2029.

"Compared with orbital computing, ground-based computing power is simply negligible," he wrote.

"Orbital computing power" refers to computing capacity such as AI and GPUs deployed in satellites or space data centers. Potential advantages include space solar power supply, heat dissipation conditions, and the ability to launch and deploy at scale. As demand for AI training and inference surges, ground-based data centers are facing multiple constraints such as power shortages, slow grid expansion, and tight land and water resources.

Supporters believe that sending servers, GPUs, and storage devices into low Earth orbit can take advantage of stronger and more continuous sunlight in space and break free from dependence on the ground power grid. Musk's latest statement provides an ambitious vision for SpaceX's expansion into AI infrastructure, as its xAI business is competing with Microsoft, Google, Amazon, and OpenAI.

Musk's post above was in fact a reply to a financial forecast made by venture capitalist Shaun Maguire. He estimated that SpaceX's revenue could soar from $51.5 billion in 2026 to $274 billion in 2027, and reach $731 billion in 2028.

For investors, these forecasts highlight the huge potential revenue opportunities in the AI infrastructure sector, while also revealing the capital investment and execution risks required to build that infrastructure. Maguire's model assumes that SpaceX's ground computing capacity will grow from 1 gigawatt in early 2026 to 21 gigawatts by the end of 2028, with each gigawatt generating about $40 billion in revenue per year.

The model predicts that EBITDA will reach $536 billion in 2028, but because annual capital expenditure is expected to reach $369 billion, free cash flow will not turn positive until the middle of that year.

The Sequoia Capital partner cautioned that his forecast carries considerable uncertainty due to potential political obstacles and supply chain constraints. Sequoia Capital holds a large stake in SpaceX. However, Musk believes these forecasts may be too conservative, meaning that even investors holding shares in the company may be underestimating its growth potential.

It is reported that SpaceX has applied to build an orbital data center constellation of up to 1 million satellites, operating at altitudes between 500 and 2,000 kilometers, and plans to transmit data in coordination with the Starlink network. And SpaceX is not the only player in this track.

For example, Google's Suncatcher project sent its first experimental spacecraft into orbit in October to explore the feasibility of running machine learning computing power in space; Blue Origin, backed by Bezos, also submitted a plan called Project Sunrise to the U.S. Federal Communications Commission, planning to deploy up to 51,600 data center satellites.

On the other hand, Musk also questioned outside speculation about the cost of the company's AI infrastructure buildout. Kyle Reidhead, an analyst and CEO of the investment newsletter "Milk Road," earlier made a forecast that SpaceX may spend $470 billion on capital expenditure over the next five quarters and accumulate about $150 billion in debt.

In response, Musk said our capital efficiency has been "seriously underestimated."

(Source: Cailian Press) Middle of article body Embedded advertisement