SpaceX and Nvidia have turned what was once a speculative idea into a concrete roadmap: a network of AI data centers operating in low Earth orbit. The first satellites are slated for launch in the fourth quarter of 2027, with the system expected to reach meaningful operational capacity by 2028. This move comes on the heels of SpaceX's record-breaking public listing in June 2026, which gave the company fresh capital and a higher public profile.
Why put data centers in space?
The pitch is straightforward: space offers a unique combination of abundant solar energy, a vacuum for cooling, and a vantage point that makes it harder for any single government or adversary to disrupt. For a world increasingly dependent on AI, the ability to process data beyond national borders could be a game-changer. Each satellite is designed to be a dense, cost-effective alternative to ground-based facilities, with a smaller physical footprint and higher energy efficiency.
Nvidia will be the exclusive hardware provider, supplying its Vera CPU architecture to power the orbital network. This is a significant expansion of Nvidia's data center business, which has traditionally been rooted in terrestrial facilities. The partnership also opens a new demand driver for Nvidia's chips, a point that has not been lost on Wall Street analysts, who remain bullish on the semiconductor giant.
Regulatory and financial backing
SpaceX has already begun the regulatory process, seeking approval for a constellation of up to one million specialized orbital nodes. That's a massive undertaking, but the company has a financial engine to fuel it: Starlink. In 2025 alone, Starlink generated over $11 billion in revenue, providing the steady cash flow needed to fund such ambitious R&D without straining SpaceX's core operations.
The plan hinges on the continued success of the Starship rocket system, which recently completed its fourteenth test flight, demonstrating improved reliability. Reusability is key—by lowering launch costs, Starship makes it economically feasible to deploy thousands of computing satellites in a relatively short time.
This is not just a tech story; it's a story about how infrastructure is evolving. As terrestrial data centers face growing scrutiny over energy use and environmental impact—see California's new data center rules and Texas halting data center permits—the idea of moving some of that load to orbit becomes more attractive. It also raises questions about who controls the internet's backbone, a topic that resonates across the Americas, where connectivity and digital sovereignty are ongoing debates.
For Latino communities, both in the U.S. and abroad, this development could have ripple effects. Cheaper, more resilient computing might eventually lower the cost of digital services in regions that currently rely on expensive terrestrial infrastructure. But it also underscores the need for equitable access to the benefits of AI, a conversation that is already happening in places like Venezuelan diaspora communities and beyond.
The timeline is ambitious, and technical hurdles remain. But with SpaceX's track record and Nvidia's hardware expertise, the prospect of orbital data centers is no longer science fiction. It's a plan with a date.
