What Telecoms Can Teach the Space Data Centre Race

This Industry Viewpoint was authored by Tae Oh, founder of Spacecoin.

The space data centre race is moving faster than the rules needed to govern it. Google’s Project Suncatcher, Starcloud’s GPU satellites, China’s Three-Body Computing Constellation – each is a serious bet that the future of AI infrastructure lies beyond Earth. The logic makes sense: space offers constant solar power, natural cooling, and escape from the land and energy limits strangling data centres on the ground. But the way these systems are being built should give pause. Every major player is constructing its own stack – satellites, inter-satellite links, ground stations, backhaul. These networks don’t connect to each other. They’re not designed to. The companies building them have the money to go it alone, so they are.

Telecoms veterans have seen this before. Before interconnection agreements became standard, calling between phone networks was either impossible or required expensive manual switching. AT&T and independent carriers operated as separate worlds for decades. In mobile, the early 1990s fractured into incompatible standards – GSM in Europe, CDMA in the US, PDC in Japan. A phone that worked in London was useless in New York. Roaming meant bilateral deals negotiated country by country, carrier by carrier. It took years of regulatory pressure and commercial pain before the industry accepted a basic truth: cooperation on infrastructure makes competition on services possible.

Telecoms veterans have seen this before. Before interconnection agreements became standard, calling between phone networks was either impossible or required expensive manual switching. AT&T and independent carriers operated as separate worlds for decades. In mobile, the early 1990s fractured into incompatible standards – GSM in Europe, CDMA in the US, PDC in Japan. A phone that worked in London was useless in New York. Roaming meant bilateral deals negotiated country by country, carrier by carrier. It took years of regulatory pressure and commercial pain before the industry accepted a basic truth: cooperation on infrastructure makes competition on services possible.

Eventually, economics forced the issue. The cost of going it alone simply didn’t add up. Today, undersea cables are built by consortiums sharing construction costs. Mobile towers are increasingly owned by independent companies that operators pay to access rather than building duplicate steel and concrete. Internet exchange points let rival networks hand off traffic efficiently. These arrangements emerged because duplicating physical infrastructure was financial suicide.

Space infrastructure faces the same economics, just with bigger numbers. A meaningful satellite constellation likely requires hundreds of millions to billions in upfront investment. Building three or four parallel systems – with no ability to share capacity or route around outages – means either massive consolidation down the line or a lot of expensive failures. The model that made global telecommunications work was shared infrastructure: common standards for authentication and handoffs, neutral interconnection points. Applied to space, this would let multiple operators deploy computing without each building everything from scratch.

Some companies are already building toward this. Instead of closed private networks, they’re creating mesh networks of relay satellites that any operator can plug into – shared pipes in orbit, essentially. The access rules are baked into the technology: meet the technical standards and you’re in, no need to negotiate deals with whoever built it. Data moves between operators and settlement happens automatically. For telecoms, this should sound familiar: it’s how roaming and interconnection work today, just designed in from day one rather than bolted on after years of friction.

Space data centres won’t exist in isolation. They’ll need ground stations, fibre backhaul, spectrum coordination, connections to terrestrial networks – all things telecoms already provides. If orbital infrastructure develops with open standards, there’s a natural role for telecoms expertise and assets. The alternative – a landscape where a handful of tech giants build everything from satellite to end user – leaves little room for outside partners. The space data centre boom is just beginning. How it gets built will shape who participates and who watches from the sidelines.

If you haven't already, please take our Reader Survey! Just 3 questions to help us better understand who is reading Telecom Ramblings so we can serve you better!

Categories: Industry Viewpoint · satellite