{"paper":{"title":"Toward Communication-Efficient Space Data Centers: Bottlenecks, Architectures, and New Paradigms","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Space data centers reduce uplink pressure by transmitting compact semantic representations instead of raw data for AI tasks.","cross_cats":[],"primary_cat":"cs.NI","authors_text":"Jinbo Hou, Kezhi Wang, Minghao Sun, Xiaoli Chu, Zehui Chen","submitted_at":"2026-05-12T19:32:08Z","abstract_excerpt":"The rapid growth of foundation model training and large-scale AI services has driven ground data centers toward unprecedented power densities, intensifying challenges in energy supply, cooling, and spatial scalability. Space Data Centers (SDCs) have emerged as a promising paradigm for hosting energy-intensive computing infrastructures in orbit, leveraging continuous solar energy and radiative cooling advantages. However, unlike ground facilities primarily constrained by power and site availability, SDCs are fundamentally limited by communication capability. The gap between petabit-scale intern"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"By transmitting compact, task-relevant semantic representations instead of raw data, uplink pressure can be substantially reduced. The feasibility of communication-efficient orbital AI infrastructures is demonstrated through the evaluation of a multi-layer heterogeneous SDC framework consisting of relay satellites and orbital computing nodes operating under coupled energy and thermal constraints.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"That semantic representations can be generated and transmitted without losing critical task-relevant information needed for foundation model training and large-scale AI services, while the multi-layer framework can operate feasibly under real coupled energy and thermal constraints.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"Semantic communication in a multi-layer heterogeneous space data center framework can substantially reduce uplink pressure for orbital AI by sending compact representations rather than raw data.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Space data centers reduce uplink pressure by transmitting compact semantic representations instead of raw data for AI tasks.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"255ed6693f5980cdc0c34d734ad6e8fb032ee70d36c86f50bb0e2d8d6da3c65d"},"source":{"id":"2605.12681","kind":"arxiv","version":1},"verdict":{"id":"b7b45972-0f87-446d-a3d7-ff112772719b","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-14T19:45:59.329919Z","strongest_claim":"By transmitting compact, task-relevant semantic representations instead of raw data, uplink pressure can be substantially reduced. The feasibility of communication-efficient orbital AI infrastructures is demonstrated through the evaluation of a multi-layer heterogeneous SDC framework consisting of relay satellites and orbital computing nodes operating under coupled energy and thermal constraints.","one_line_summary":"Semantic communication in a multi-layer heterogeneous space data center framework can substantially reduce uplink pressure for orbital AI by sending compact representations rather than raw data.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"That semantic representations can be generated and transmitted without losing critical task-relevant information needed for foundation model training and large-scale AI services, while the multi-layer framework can operate feasibly under real coupled energy and thermal constraints.","pith_extraction_headline":"Space data centers reduce uplink pressure by transmitting compact semantic representations instead of raw data for AI tasks."},"references":{"count":15,"sample":[{"doi":"","year":2025,"title":"Ag¨ uera y Arcas, T","work_id":"0ce13201-8f1b-4483-9391-b2e951f9d953","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Why we should train ai in space,","work_id":"b33c6bbc-fe20-479b-8173-a69f4f9527fa","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2025,"title":"Towards space-based computing infrastructure net- work: Development trends, network architecture, challenges analysis, and key technologies","work_id":"17319b88-a902-4d4f-b1c8-37fa6d0e8650","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2010,"title":"Analysis and design of a solar rectenna,","work_id":"35f8df7a-f1a9-4820-9f1d-ba2da7c37a39","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"2024 united states data center energy usage report,","work_id":"3c6ff266-22a6-4f79-a510-12d1089d7105","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":15,"snapshot_sha256":"dce0e03c3699900f758826b893ed868843c90db34652e0c7842d3bb864036f4c","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"}