space datacenter network

This work addresses the challenges of constructing high-density distributed space-based data centers in low Earth orbit (LEO) by proposing two parametric satellite constellation architectures—planar and three-dimensional—that optimize geometric layouts under constraints including minimum inter-satellite distance, unobstructed solar power access, and stable inter-satellite links. The design maps a VL2-inspired Clos network topology onto feasible inter-satellite links. Leveraging orbital dynamics modeling, numerical analysis, and integer optimization, the approach achieves densest packing in the planar configuration and enables the 3D architecture to scale satellite count proportionally to $(R_{\max}/R_{\min})^3$. Experimental results demonstrate that both architectures provide sufficient persistent, obstruction-free links to replicate terrestrial data center switching fabrics, while quantifying the trade-off between per-satellite link capacity and the number of dedicated switching satellites.

This study addresses the limitations of traditional connection-centric satellite communications, which struggle to meet the mission-oriented, service-driven information exchange demands of future data-intensive and AI-enabled space applications. To bridge this gap, the paper proposes a novel hierarchical Service-oriented Space Data Center (SDC) architecture, redefining orbital nodes as intelligent service entities that integrate communication, computation, storage, and control capabilities rather than serving merely as relays. The design encompasses coordinated deployment strategies and application scenarios across four functional layers: access, relay, computing, and control. By jointly optimizing on-orbit computing resource scheduling and low-latency control mechanisms, the proposed SDC architecture significantly reduces control-plane latency, as demonstrated through simulations, thereby effectively supporting next-generation intelligent space applications.

This work addresses the challenge of constrained data transmission for large-scale AI tasks in space due to limited ground-to-orbit communication bandwidth. To overcome this bottleneck, the study proposes a novel orbital data center architecture that integrates semantic communication with a multi-layer heterogeneous satellite network comprising relay and in-orbit computing nodes. By transmitting semantic information instead of raw data, the approach substantially reduces uplink traffic. A coupled energy–thermal management model is introduced to evaluate system feasibility. The first comprehensive analysis demonstrates that, under gigabit-class ground-to-orbit links, the proposed architecture can support petabyte-scale internal data exchange, thereby establishing a foundation for scalable and energy-efficient orbital AI systems.

This work addresses the challenges posed by the rapid proliferation of low Earth orbit satellites—including high data downlink costs, link congestion, significant latency, and ground station scheduling constraints—by proposing, for the first time, an on-orbit AI-driven multi-tenant Space Data Center (SDC) constellation architecture that brings computation and intelligent processing capabilities to the space edge. The proposed architecture integrates orbital design, inter-satellite links, distributed resource management, and AI service orchestration to enable on-demand access for multiple users. Validation through Earth observation and lunar exploration scenarios demonstrates the SDC’s technical feasibility and economic advantages, offering a novel paradigm for future space-based intelligent infrastructure.

This work addresses the challenges of mobility management, interference control, and spectral efficiency in large-scale direct-to-cell low Earth orbit (LEO) satellite communications by proposing a novel multi-orbit hierarchical space data center architecture that integrates LEO, medium Earth orbit (MEO), and geostationary Earth orbit (GEO) segments. The architecture uniquely combines distributed in-orbit computing, energy-aware scheduling, AI-driven hierarchical control, and computation-aware routing to enable synergistic optimization across access, regional aggregation, and global coordination layers. By transcending conventional relay paradigms, the resulting intelligent constellation system significantly enhances scalability and robustness, establishing a new architectural paradigm for 6G non-terrestrial networks capable of efficiently supporting massive direct terminal connectivity and intelligent services.

This study evaluates the feasibility of large-scale AI data centers in low Earth orbit as an alternative to terrestrial facilities. Accounting for key differences between orbital and ground-based systems—including launch costs, power supply, thermal dissipation, radiation exposure, and atmospheric re-entry—it proposes a mesh network architecture based on laser inter-satellite links. The work introduces, for the first time, dual-cut bandwidth, dual-cut strength, and roofline models to systematically assess the performance of space-based AI computing. Comparative analysis with Clos networks reveals fundamental limitations of the orbital environment for training large language models: while orbital data centers can support AI inference tasks, they fall significantly short of ground-based counterparts in terms of cost efficiency and scalability for training state-of-the-art foundation models.

Serverless computing has matured into an effective execution model for edge cloud environments, enabling function level decomposition, demand driven scaling, and workflow execution across stable, well provisioned infrastructure. This success motivates extending it to the edge cloud space continuum, where Low Earth Orbit (LEO) constellations are increasingly explored as distributed compute substrates. However, existing serverless orchestration is not directly applicable in this setting, where LEO systems impose time varying contact graphs, intermittent link availability, and strict feasibility constraints on energy, memory, communication, and operational cost. This article identifies ten broken assumptions in existing serverless orchestration and organizes them into three core challenges: spatiotemporal execution over dynamic graphs, constraint aware function placement and scaling, and correctness and progress under decentralized and delayed state. It then proposes an architecture that enables robust and efficient serverless execution across the continuum, grounded in these challenges and demonstrated through a representative flood response use case.

This work addresses the lack of systematic educational resources in high-performance computing (HPC) networking, which poses a significant barrier for researchers entering the field. It presents the first comprehensive integration of the HPC networking stack, covering communication protocol layers, programming interfaces such as MPI, control plane mechanisms, high-speed interconnect technologies, and custom link-layer hardware. The exposition is anchored by a detailed case study of the El Capitan supercomputer architecture at Lawrence Livermore National Laboratory. By offering a well-structured, practice-oriented primer, this contribution fills a critical educational gap and substantially lowers the entry barrier for researchers seeking to master core HPC networking technologies.

This work addresses the lack of a high-throughput, general-purpose solution for large-scale AI training that simultaneously respects physical constraints and optimizes network topology, routing, and collective communication. The authors propose TONS, a framework that enables automated throughput-optimized network synthesis for AI supercomputers. TONS formulates topology synthesis as a linear optimization problem and scales to thousands of nodes by integrating theoretical insights with heuristic methods. It also introduces a deadlock-free routing mechanism supporting limited virtual channels and fault tolerance in optical switching. Under realistic deployment constraints, TONS achieves geometric mean speedups of 2.1× and 1.6× over the best TPU v4/5p torus variants for uniform random and all-to-all communication patterns, respectively.