Compute Dependence and Infrastructure Autonomy in Small and Medium States

算力依赖风险与中小国家数字基础设施自主性 Not peer reviewed. Chinese policy-analysis manuscript with an English abstract and reproducible illustrative cost calculations. This is not a national field experiment, country-level dataset or causal policy evaluation. The paper examines compute dependence as a conditional risk involving service continuity, exit capability, operational knowledge and governance. It does not establish inevitable colonial domination or make an allegation against any named country or supplier. Ten scholarly and official sources are identified in a claim-scope register. Statistics on data-centre capacity, cloud-region availability, research spending and electricity demand are not treated as interchangeable measures of AI compute. The original deterministic model enumerates 324 explicitly assumed combinations of annual fixed cost, cloud price, incremental local cost, utilization and relative task efficiency. In an illustrative baseline, reducing task efficiency from 1 to 0.75 raises break-even utilization from 53.51% to 77.83%. These inputs are not observed market quotes or calibrated national costs. Grid frequencies are not empirical probabilities. The paper proposes critical-service continuity capacity, regionally governed shared resources, portable commercial cloud access and domestic operational competence, with limitations and a prospective evaluation design. The PDF and editable DOCX represent the same paper. The ZIP contains the original Python reproduction script, assumed scenario grid, results, source register, README and checksums. No raw personal data, email credentials or third-party full texts are included. CC BY 4.0 covers the original article, code and original scenario outputs; it does not relicense the cited third-party sources. Author: YANGZHOU. Affiliation: Genbibao Embodied AI Research Institute, Shenzhen, China. ORCID: https://orcid.org/0009-0005-4736-2097 . Correspondence: [email protected] . The correspondence address does not establish employment at Sichuan University. Funding: The author confirms no external research funding for this study. Competing interests: Institute affiliation and marketing-related business participation are disclosed; the author confirms no additional interests. OpenAI Codex assisted source discovery, evidence organization, model and code development, execution, checks and drafting. AI is not an author. No human-participant research was conducted. Disclosure clarification added on 20 November 2026: The author's supplied biography describes past activities involving QiSense and an intelligent computing-centre project. The author has confirmed that there is no continuing association with those activities. The historical biographical description is not treated as evidence validating the paper's model or policy conclusions. This manuscript is a policy analysis and is not an independent procurement assessment.

Authors

Publication Details

Journal
Zenodo (CERN European Organization for Nuclear Research)
Published
2026-11-20
DOI
https://doi.org/10.5281/zenodo.24376019
Primary Topic
Security, Politics, and Digital Transformation
Type
preprint
Controls
|||
ALL TIME
JAN
FEB
MAR
APR
MAY
JUN
JUL
AUG
SEP
OCT
NOV
preprint

Compute Dependence and Infrastructure Autonomy in Small and Medium States

YANGZHOU
Zenodo (CERN European Organization for Nuclear Research)
Security, Politics, and Digital Transformation
preprint

Compute Dependence and Infrastructure Autonomy in Small and Medium States

YANGZHOU
preprint en

Abstract

算力依赖风险与中小国家数字基础设施自主性 Not peer reviewed. Chinese policy-analysis manuscript with an English abstract and reproducible illustrative cost calculations. This is not a national field experiment, country-level dataset or causal policy evaluation. The paper examines compute dependence as a conditional risk involving service continuity, exit capability, operational knowledge and governance. It does not establish inevitable colonial domination or make an allegation against any named country or supplier. Ten scholarly and official sources are identified in a claim-scope register. Statistics on data-centre capacity, cloud-region availability, research spending and electricity demand are not treated as interchangeable measures of AI compute. The original deterministic model enumerates 324 explicitly assumed combinations of annual fixed cost, cloud price, incremental local cost, utilization and relative task efficiency. In an illustrative baseline, reducing task efficiency from 1 to 0.75 raises break-even utilization from 53.51% to 77.83%. These inputs are not observed market quotes or calibrated national costs. Grid frequencies are not empirical probabilities. The paper proposes critical-service continuity capacity, regionally governed shared resources, portable commercial cloud access and domestic operational competence, with limitations and a prospective evaluation design. The PDF and editable DOCX represent the same paper. The ZIP contains the original Python reproduction script, assumed scenario grid, results, source register, README and checksums. No raw personal data, email credentials or third-party full texts are included. CC BY 4.0 covers the original article, code and original scenario outputs; it does not relicense the cited third-party sources. Author: YANGZHOU. Affiliation: Genbibao Embodied AI Research Institute, Shenzhen, China. ORCID: https://orcid.org/0009-0005-4736-2097 . Correspondence: [email protected] . The correspondence address does not establish employment at Sichuan University. Funding: The author confirms no external research funding for this study. Competing interests: Institute affiliation and marketing-related business participation are disclosed; the author confirms no additional interests. OpenAI Codex assisted source discovery, evidence organization, model and code development, execution, checks and drafting. AI is not an author. No human-participant research was conducted. Disclosure clarification added on 20 November 2026: The author's supplied biography describes past activities involving QiSense and an intelligent computing-centre project. The author has confirmed that there is no continuing association with those activities. The historical biographical description is not treated as evidence validating the paper's model or policy conclusions. This manuscript is a policy analysis and is not an independent procurement assessment.

Zenodo (CERN European Organization for Nuclear Research)
Security, Politics, and Digital Transformation
AI Navigator

Ask Laika to Summarize, Analyze, and Connect papers live on the map.

Summarize Papers & Methodologies

Extract key findings, datasets, and comparative methods across publications.

Benchmark Rankings & Visual Analytics

Rank top research institutions, authors, funders, topics, and journals by Field-Weighted Citation Impact (FWCI) and paper volume with instant charts.

Connect Distant Disciplines

Bridge topological clusters on the map to find hidden collaborative intersections.

Laika access coming soon