In the space of 10 days in March 2026, Elon Musk unveiled TeraFab, a $20-25 billion chip fabrication plant to produce radiation-hardened semiconductors for orbital data centers, and startup Starcloud closed a $170 million Series A at a $1.1 billion valuation, becoming the fastest unicorn in Y Combinator history. These announcements follow SpaceX’s January Federal Communications Commission (FCC) filing to deploy up to one million data center satellites, Google’s Project Suncatcher, and Blue Origin’s TeraWave.
This recent shift from concept to active industrial mobilization reflects two main factors:
- Growing challenges with terrestrial data center (DC) expansion.
-
Improving economics driven by the falling costs of space flight.
At Altman Solon, our data center advisory practice works with clients across the value chain — from hyperscalers and colocation operators to power developers and private equity investors. Below is our perspective on what orbital compute entails, what is required for deployment at scale, and what it means for terrestrial data center strategy.
The terrestrial breaking point
The push toward orbit is a response to very real constraints. Power interconnection queues in major data center hubs can now stretch up to seven years. EU data centers are projected to consume around 108 terawatt hours (TWh) by 2030, more than the Netherlands’ current total annual electricity consumption. Water-based cooling is driving scarcity and reputational risk. Community opposition has matured into a systematic scheduling risk. One of the industry’s biggest challenges lies in securing the land, power, water, and political license to deploy and operate facilities.
Furthermore, long-term global data center demand (measured in gigawatts) is likely to outpace supply, with power as the limiting factor. Orbital compute seeks to address this structural supply gap.
What orbital data centers offer
Orbital data centers are compute hardware hosted aboard satellites, typically in low Earth orbit (240–1,200 miles). The vision is not to transplant terrestrial hyperscale facilities directly into space, but to develop a modular, networked layer of compute satellites. Operating a data center in orbit provides several structural advantages:
However, gating issues are equally clear:
Each challenge can be addressed, but requires adding mass – for example, radiation shielding or battery back-up power, which results in more required launches and higher costs. This mass-cost spiral is the central engineering tension.
The economics: $200/kg is the critical threshold
The economic viability of space data centers depends on a set of assumptions that must hold simultaneously:
Chips in orbit can be operated until true physical end-of-life, avoiding the premature refresh cycles common in terrestrial facilities.
Launch economics could also be cross-subsidized by piggybacking on other space infrastructure, such as Starlink’s growing constellation deployment, treating compute as a marginal addition rather than a standalone launch cost.
Key finding: SpaceX’s potential ability to treat orbital data centers at least partially as a marginal addition to its existing Starlink program (combined with TeraFab’s promise of in-house chip production) creates a vertically integrated cost structure (chips, rockets, satellites, networking, captive AI demand) that would be unique among players in the industry.
Workloads that move to orbit first
Orbital compute is emerging as a new infrastructure tier; not a replacement for terrestrial data centers, but an addition. The workloads that migrate first are those where orbit’s advantages most clearly outweigh its constraints:
What this means for the value chain
Orbital compute is emerging as a new infrastructure tier; not a replacement for terrestrial data centers, but an addition. The workloads that migrate first are those where orbit’s advantages most clearly outweigh its constraints:
Possible timeline
Key finding: The key uncertainty is not technical feasibility — that has been demonstrated at least at small scale. It is launch economics and increasing scale. If Starship continues the historical trend of launch cost reduction and delivers on its cost targets, the timeline accelerates. If not, orbital compute remains a defense and space-data niche. Either way, the implications are material enough to warrant consideration today.
How Altman Solon can help
Altman Solon's data center advisory practice supports clients across the value chain on the strategic and commercial questions raised by the wider DC ecosystem and now orbital compute.
Our relevant capabilities include:
