The surge in artificial intelligence computing demand has pushed terrestrial data centers to their structural limits in power supply, cooling capacity, and site availability. The enormous electricity requirements and heat management for training large-scale AI models, combined with the burden of massive infrastructure development, are escalating rapidly. As an alternative to ease these constraints, the concept of "in-orbit data centers"—placing computing functions in Earth's orbit—is emerging as a new frontier in global technology competition.
According to Research and Markets, the orbital data center market is projected to grow from approximately $1.77 billion in 2029 to over $39 billion by 2035, representing an annual growth rate of about 67%. While still in its early stages, the sector is gaining attention as a long-term emerging market driven by power constraints and expanding AI infrastructure demand.
Dwindling Power Supply Turns Eyes to Solar Energy
The primary reason the industry is looking skyward is power. According to the U.S. Department of Energy, data centers consumed about 4.4% of total U.S. electricity in 2023, with projections suggesting this could rise to 6.7-12% by 2028 if current trends continue. Power consumption is expected to accelerate further as GPU density increases with advancing AI models.
Cooling poses another significant burden. High-performance semiconductors generate massive amounts of heat, requiring substantial cooling facilities and water resources to manage. Site selection has also become a complex challenge requiring simultaneous consideration of power grid access, cooling infrastructure, regulations, and local acceptance.
Space is evaluated as an environment relatively free from these constraints. Orbits unaffected by weather changes or day-night cycles allow stable, extended use of solar power. Radiative cooling designs utilizing the vacuum environment can minimize water usage. The virtual absence of physical site limitations also enables modular expansion of large-scale infrastructure.
U.S., China, and Europe Launch Challenges Despite Clear Technical Limitations
China currently shows the most aggressive movement. Chinese private space company ADA Space launched 12 satellites in May 2025 as the first phase of its "Three-Body Constellation" project, initiating orbital computing demonstration. The project ultimately aims to network approximately 2,800 satellites to build distributed AI computing infrastructure in space.
Each satellite reportedly carries AI processors with hundreds of tera-operations per second (TOPS) processing capability and systems capable of running models with billions of parameters. High-speed laser-based communication links between satellites enable direct data processing in orbit. The plan is to verify an efficient data processing system that transmits only necessary results rather than sending all raw data to the ground.
In the United States, Axiom Space has partnered with Amazon Web Services to install small edge computing equipment on the International Space Station in 2022, conducting demonstrations of data storage and processing in orbit. This technical trial project aims to verify whether on-site data analysis is possible in environments with limited ground communication.
In Europe, the ASCEND study led by Thales Alenia Space analyzed the technical and economic feasibility of orbital data processing infrastructure and presented a concept-level roadmap. However, these attempts remain at the proof-of-concept stage, with many assessments indicating considerable time is needed before they develop into commercial infrastructure.
Technical challenges are also clear. Space radiation can degrade semiconductor reliability, making radiation-hardened design and system redundancy essential. Heat dissipation in environments without atmosphere or fluid, inter-satellite network stability, and risks of orbital congestion and space debris collision remain problems to solve. Due to the characteristics of low-Earth-orbit satellites with limited continuous ground connectivity, many analyses suggest they are better suited for batch computing or data preprocessing rather than real-time services.
Domestic Space Data Centers Not Yet in Sight, but Supply Chain Companies Expected to Benefit
No Korean companies are directly pursuing "space data centers" yet, but the related technology ecosystem is gradually forming. In the launch vehicle sector, private space companies are challenging commercial launch services. Innospace is developing hybrid rockets for small satellite launches and pursuing commercial launch contracts with overseas satellite customers, attempting to enter the private launch market. Perigee Aerospace is also pursuing small orbital launch vehicle development to expand Korea's private launch capabilities.
In the mid-to-long term, if demand for space-based data processing and satellite utilization expands, domestic industries supplying related technologies are more likely to benefit before platform builders. In the launch vehicle sector, private companies preparing small satellite launch services and aerospace manufacturers participating in rocket components and engine production are expected to see direct demand growth from increased satellite launches.
The satellite manufacturing industry designing and building satellite bodies and payloads also has significant growth potential alongside expansion of satellite constellations for Earth observation, communications, and data collection. In semiconductors, capabilities in low-power, high-reliability design and radiation-hardening technology could open new markets for space-environment electronic components. Network, optical, and electronics companies handling inter-satellite communications, ground station operations, data processing software, and power and thermal management equipment are also expected to see related demand as space infrastructure expands.