Track papers across labs, topics, and trends.
Search papers, labs, and topics across Lattice.
This paper assesses the viability of deploying large-scale AI data centers in low-Earth orbit (LEO) as a potential cost-effective alternative to terrestrial facilities by analyzing various factors such as launch costs, power generation, cooling, radiation exposure, and compute-network performance. The authors highlight a significant shift from terrestrial Clos networks to space-based mesh networks utilizing laser inter-satellite links, which impacts network performance. Ultimately, while LEO-based inference shows promise, the study concludes that training frontier-scale large language models (LLMs) in orbit is unlikely to compete with existing terrestrial data centers due to performance limitations.
LEO-based AI data centers may excel at inference, but training frontier-scale LLMs in orbit falls short against terrestrial counterparts.
This paper evaluates whether large-scale AI data centers deployed in low-Earth orbit (LEO) could become a cost-effective alternative to terrestrial facilities. The analysis compares orbital and ground-based systems across launch cost, power generation, cooling, radiation exposure, and atmospheric reentry, as well as compute-network performance. A key distinction is the shift from terrestrial Clos networks to space-based mesh networks using laser inter-satellite links. Using bisection bandwidth, bisection intensity, and roofline-style models, we show that while LEO-based inference may be feasible, training frontier-scale LLMs in orbit is unlikely to be competitive with terrestrial data centers.