A Pragmatic Path Toward Orbital Data Centers

JASON ASPIOTIS OF AXIOM SPACE OUTLINES THE ECONOMIC CASE FOR SPACE-BASED CLOUD COMPUTING AT SPACENEXT 2026

At spaceNEXT 2026, Jason Aspiotis of Axiom Space offered a provocative thesis about one of the most talked-about emerging ideas in the commercial space industry: orbital data centers. While the concept of moving computing infrastructure into space often sparks debates about engineering challenges, Aspiotis argued that the real obstacle isn’t physics at all.

“Orbital data centers are not a physics problem,” he told the audience. “They’re an economics problem.”

The idea behind orbital data centers—often shortened to ODCs—is straightforward but ambitious: satellites designed specifically to perform data storage, processing, and AI workloads in orbit. Instead of sending vast quantities of raw satellite data back to Earth for analysis, these platforms would process information directly in space and transmit only the insights that matter.

For Aspiotis and his team at Axiom Space, this vision is no longer theoretical. Axiom is among a small number of companies already testing the concept through prototype systems operating in orbit, including hardware currently flying aboard the International Space Station and on free-flying platforms developed in collaboration with partners such as Kepler. These early systems are designed to demonstrate what space-based cloud computing could look like in practice.

At its core, Aspiotis explained, an orbital data center is essentially a purpose-built computing satellite—a platform capable of connecting to other spacecraft, processing their data, and delivering insights either to other satellites or back to Earth through radio-frequency and optical communications.

The need for such infrastructure is growing rapidly. Over the past decade, the number of satellites in orbit has increased dramatically, particularly in areas such as Earth observation, synthetic aperture radar, and national security monitoring. Each of these platforms generates enormous volumes of data—often measured in terabytes or even petabytes.

“The value proposition is about speed to insights,” Aspiotis said. “It’s about making decisions in space as fast as possible.”

By processing data in orbit, satellite operators could dramatically shorten the time between collecting information and acting on it. For defense agencies, environmental monitoring systems, and commercial analytics companies, that speed could be a major advantage.

Beyond data processing, orbital data centers could also strengthen the resilience and security of space infrastructure. As global economies become increasingly dependent on satellite networks for communications, navigation, and sensing, distributing computing capabilities across multiple platforms in orbit could help reduce vulnerabilities and improve redundancy.

Another potential advantage is the ability to create sovereign cloud environments in space. In this model, orbital infrastructure could provide secure computing environments accessible from anywhere on Earth, allowing governments or enterprises to operate independent cloud systems without relying solely on terrestrial data centers.

While those near-term use cases focus on supporting satellite systems, Aspiotis pointed to a much larger long-term vision: relocating portions of Earth’s growing digital infrastructure into orbit. As artificial intelligence workloads expand and global demand for computing continues to surge, terrestrial data centers are increasingly constrained by power consumption, cooling requirements, and land availability.

Space offers a different set of conditions. Solar energy is abundant, cooling is naturally efficient in the vacuum of space, and orbital infrastructure avoids the land-use conflicts that often accompany large terrestrial facilities.

“We’re not building data centers to replace schools or forests or hospitals,” Aspiotis said. “We’re building them in space, where real estate is effectively infinite.”

The market opportunity for orbital computing is expected to unfold gradually. Early adoption will likely come from government agencies and satellite operators seeking faster access to data analytics in orbit. Industry estimates suggest that this initial market could reach several billion dollars in value.

As adoption expands across civil and commercial sectors, analysts project the market for orbital computing services could reach roughly $25 billion by 2035. Beyond that, if space-based infrastructure begins supporting large-scale AI workloads and cloud computing services, the opportunity could grow even larger—potentially reaching trillion-dollar scale over time.

That trajectory would represent a major milestone for the space industry, which has long searched for a commercial application capable of transforming the sector’s economics.

“This could be the killer app the space industry has been looking for,” Aspiotis said.

Yet despite the excitement surrounding the concept, significant economic challenges remain. Running computing workloads in orbit is still dramatically more expensive than performing the same tasks on Earth. Aspiotis estimated that launching and operating orbital computing systems today could make space-based cloud services roughly twenty times more expensive than terrestrial alternatives.

However, he emphasized that the economics of space infrastructure have changed rapidly in recent years. Launch costs have fallen dramatically—by as much as fifty-fold over the past two decades—and continued advances in reusable rockets and large satellite constellations are expected to drive costs even lower.

As early adopters begin paying for orbital computing services, those investments could help accelerate the scale needed to bring costs down. Over time, Aspiotis believes, orbital computing capacity could expand from experimental systems measured in kilowatts to large-scale infrastructure measured in megawatts—and eventually gigawatts.

Such growth will depend on a broader ecosystem of technologies that are already emerging in the space industry. Satellite mesh networks are expanding rapidly, enabling spacecraft to communicate directly with one another. In-space servicing and robotic assembly systems are also advancing, making it possible to repair, refuel, or upgrade spacecraft after launch.

These capabilities could prove especially important for orbital data centers, whose computing hardware may need to be replaced or upgraded every few years to keep pace with rapid advances in processor technology.

Government policy may also play a role in shaping the market. Aspiotis noted that public investment and procurement programs have historically helped catalyze major space infrastructure developments. Similar approaches—such as long-term commitments to purchase orbital computing services—could help create the stable demand needed to support early commercial deployments.

Despite some ambitious projections circulating in the industry, Aspiotis advocated for a measured, pragmatic path forward. While some proposals envision massive constellations of AI-focused satellites in the near future, he believes the industry will scale more gradually as technologies mature and markets develop.

Within the next decade or two, however, he believes orbital computing capacity could grow to gigawatt-scale infrastructure in space. If that happens, the implications would extend far beyond cloud computing.

“If we can build gigawatts of compute in space,” Aspiotis said, “we can build anything.”

For Aspiotis, orbital data centers represent more than just a new technology platform. They could become foundational infrastructure for a much larger transformation—one in which humanity builds increasingly complex systems beyond Earth.

And if that future unfolds, the cloud may not just be above us metaphorically.

It may literally be in orbit.