智算中心论文专站

AIDC Research Papers

Liquid Cooling AI Data Center Power & Thermal Systems
Current Issue

Volume 2026 · Issue 06-06

按期刊卷期页方式整理本期论文。每条仅使用日报已列出的可追溯公开来源,不新增未经核验事实。

Research Article热管理与液冷

XWind: A Cross-site Router for Large Language Model Inference Serving at Renewable Energy Farms

Tella Rajashekhar Reddy、Atharva Deshmukh、Liangcheng Yu、Chaojie Zhang、Mike Shepperd、Rohan Gandhi、Anjaly Parayil、Srinivasan Iyengar

Published 2026-05-22 · arXiv · Credibility S

AI

Abstract, interpretation and reference

Abstract

AI

中文解读

背景:AI 数据中心负载、功率密度和能源约束同步上升,液冷、热管理和数据中心能效正在成为智算中心设计的关键变量。问题:论文聚焦现有方案在效率、可靠性或工程协同上的瓶颈。方法:摘要显示作者采用仿真建模和情景分析,把运行负载、冷却/能源系统和基础设施约束放在同一分析框架中。结果:研究重点指向冷却效率、能源利用或运维策略的改进方向。意义:对日报读者而言,它可用于判断液冷方案、热管理路线和高密度部署节奏。仍需结合全文实验条件、样本范围和成本假设核验。

参考文献

Tella Rajashekhar Reddy, Atharva Deshmukh, Liangcheng Yu, 等. XWind: A Cross-site Router for Large Language Model Inference Serving at Renewable Energy Farms[J/OL]. (2026-05-22)[2026-06-06]. https://arxiv.org/abs/2605.23348.

Full text
热管理与液冷 论文图示
Research Article热管理与液冷

PALS: Power-Aware LLM Serving for Mixture-of-Experts Models

Can Hankendi、Rana Shahout、Minlan Yu、Ayse K. Coskun

Published 2026-05-20 · arXiv · Credibility S

Large language model (LLM) inference has become a dominant workload in modern data

Abstract, interpretation and reference

Abstract

Large language model (LLM) inference has become a dominant workload in modern data

中文解读

背景:AI 数据中心负载、功率密度和能源约束同步上升,液冷、热管理和数据中心能效正在成为智算中心设计的关键变量。问题:论文聚焦现有方案在效率、可靠性或工程协同上的瓶颈。方法:摘要显示作者采用仿真建模和情景分析,把运行负载、冷却/能源系统和基础设施约束放在同一分析框架中。结果:研究重点指向跨地域数据中心负载与电力资源之间的调度关系。意义:对日报读者而言,它可用于判断液冷方案、热管理路线和高密度部署节奏。仍需结合全文实验条件、样本范围和成本假设核验。

参考文献

Can Hankendi, Rana Shahout, Minlan Yu, 等. PALS: Power-Aware LLM Serving for Mixture-of-Experts Models[J/OL]. (2026-05-20)[2026-06-06]. https://arxiv.org/abs/2605.21427.

Full text
热管理与液冷 论文图示
Research Article热管理与液冷

Sustainability in Telecom: Energy -Efficient Networks and Circular Economy Models to Reduce Carbon Footprints and Increase Efficiency

Praveen Hegde、Robin Joseph Varughese

Published 2026-05-15 · arXiv · Credibility S

The increasing environmental impact of the telecom industry has heightened the need for sustainable telecommunications networks. With skyrocketing data

Abstract, interpretation and reference

Abstract

The increasing environmental impact of the telecom industry has heightened the need for sustainable telecommunications networks. With skyrocketing data

中文解读

背景:AI 数据中心负载、功率密度和能源约束同步上升,液冷、热管理和数据中心能效正在成为智算中心设计的关键变量。问题:论文聚焦现有方案在效率、可靠性或工程协同上的瓶颈。方法:摘要显示作者采用仿真建模和情景分析,把运行负载、冷却/能源系统和基础设施约束放在同一分析框架中。结果:研究重点指向冷却效率、能源利用或运维策略的改进方向。意义:对日报读者而言,它可用于判断液冷方案、热管理路线和高密度部署节奏。仍需结合全文实验条件、样本范围和成本假设核验。

参考文献

Praveen Hegde, Robin Joseph Varughese. Sustainability in Telecom: Energy -Efficient Networks and Circular Economy Models to Reduce Carbon Footprints and Increase Efficiency[J/OL]. (2026-05-15)[2026-06-06]. https://arxiv.org/abs/2605.16109.

Full text
热管理与液冷 论文图示
Research Article热管理与液冷

Flexibility-Aware Framework for Efficient Planner-Initiated Siting of Data Center

Dongjoo Kim、Lin Dong、Le Xie

Published 2026-05-14 · arXiv · Credibility S

Explosive growth in energy

Abstract, interpretation and reference

Abstract

Explosive growth in energy

中文解读

背景:AI 数据中心负载、功率密度和能源约束同步上升,液冷、热管理和数据中心能效正在成为智算中心设计的关键变量。问题:论文聚焦现有方案在效率、可靠性或工程协同上的瓶颈。方法:摘要显示作者采用框架构建和频域/系统级分析,把运行负载、冷却/能源系统和基础设施约束放在同一分析框架中。结果:研究重点指向冷却效率、能源利用或运维策略的改进方向。意义:对日报读者而言,它可用于判断液冷方案、热管理路线和高密度部署节奏。仍需结合全文实验条件、样本范围和成本假设核验。

参考文献

Dongjoo Kim, Lin Dong, Le Xie. Flexibility-Aware Framework for Efficient Planner-Initiated Siting of Data Center[J/OL]. (2026-05-14)[2026-06-06]. https://arxiv.org/abs/2605.14714.

Full text
热管理与液冷 论文图示
Research Article热管理与液冷

Battery-Assisted Operation of Hyperscale AI Data Centers under Connect-and-Manage Interconnection Practices

Xin Lu、Jing Qiu、Jiafeng Lin、Sihai An、Mingyang Sun、Junhua Zhao

Published 2026-05-13 · arXiv · Credibility S

Emerging connect-and-manage practices allow new transmission-connected mega-loads to connect while enforcing time-varying admissible power exchange limits at the point of common coupling (PCC) in real time. Hyperscale artificial intelligence data

Abstract, interpretation and reference

Abstract

Emerging connect-and-manage practices allow new transmission-connected mega-loads to connect while enforcing time-varying admissible power exchange limits at the point of common coupling (PCC) in real time. Hyperscale artificial intelligence data

中文解读

背景:AI 数据中心负载、功率密度和能源约束同步上升,液冷、热管理和数据中心能效正在成为智算中心设计的关键变量。问题:论文聚焦现有方案在效率、可靠性或工程协同上的瓶颈。方法:摘要显示作者采用文献摘要中的模型、实验或案例分析,把运行负载、冷却/能源系统和基础设施约束放在同一分析框架中。结果:研究重点指向冷却效率、能源利用或运维策略的改进方向。意义:对日报读者而言,它可用于判断液冷方案、热管理路线和高密度部署节奏。仍需结合全文实验条件、样本范围和成本假设核验。

参考文献

Xin Lu, Jing Qiu, Jiafeng Lin, 等. Battery-Assisted Operation of Hyperscale AI Data Centers under Connect-and-Manage Interconnection Practices[J/OL]. (2026-05-13)[2026-06-06]. https://arxiv.org/abs/2605.14105.

Full text
热管理与液冷 论文图示
Research Article热管理与液冷

Toward Communication-Efficient Space Data Centers : Bottlenecks, Architectures, and New Paradigms

Minghao Sun、Zehui Chen、Jinbo Hou、Kezhi Wang、Xiaoli Chu

Published 2026-05-12 · arXiv · Credibility S

The rapid growth of foundation model training and large-scale AI

Abstract, interpretation and reference

Abstract

The rapid growth of foundation model training and large-scale AI

中文解读

背景:AI 数据中心负载、功率密度和能源约束同步上升,液冷、热管理和数据中心能效正在成为智算中心设计的关键变量。问题:论文聚焦现有方案在效率、可靠性或工程协同上的瓶颈。方法:摘要显示作者采用仿真建模和情景分析,把运行负载、冷却/能源系统和基础设施约束放在同一分析框架中。结果:研究重点指向冷却效率、能源利用或运维策略的改进方向。意义:对日报读者而言,它可用于判断液冷方案、热管理路线和高密度部署节奏。仍需结合全文实验条件、样本范围和成本假设核验。

参考文献

Minghao Sun, Zehui Chen, Jinbo Hou, 等. Toward Communication-Efficient Space Data Centers : Bottlenecks, Architectures, and New Paradigms[J/OL]. (2026-05-12)[2026-06-06]. https://arxiv.org/abs/2605.12681.

Full text
热管理与液冷 论文图示
Research Article算电协同

Carbon-Aware Compute--Power Scheduling for AI Data Centers with Microgrid Prosumer Operations

Johnny R. Zhang、Gaoyuan Du、Qianyi Sun、Shiqi Wang、Jiaxuan Li、Xian Sun

Published 2026-05-12 · arXiv · Credibility S

AI

Abstract, interpretation and reference

Abstract

AI

中文解读

背景:AI 数据中心负载、功率密度和能源约束同步上升,算力负载与电网侧资源的协同调度正在成为智算中心设计的关键变量。问题:论文聚焦现有方案在效率、可靠性或工程协同上的瓶颈。方法:摘要显示作者采用建模优化、调度分析或算法评估,把运行负载、冷却/能源系统和基础设施约束放在同一分析框架中。结果:研究重点指向AI 负载波动对电网设备寿命和调频边界的影响。意义:对日报读者而言,它可用于判断智算中心建设是否受电网容量、负载波动和调度机制约束。仍需结合全文实验条件、样本范围和成本假设核验。

参考文献

Johnny R. Zhang, Gaoyuan Du, Qianyi Sun, 等. Carbon-Aware Compute--Power Scheduling for AI Data Centers with Microgrid Prosumer Operations[J/OL]. (2026-05-12)[2026-06-06]. https://arxiv.org/abs/2605.03751.

Full text
算电协同 论文图示
Research Article热管理与液冷

SpaceMoE: Realizing Distributed Mixture-of-Experts Inference over Space Networks

Zhanwei Wang、Huiling Yang、Min Sheng、Khaled B. Letaief、Kaibin Huang

Published 2026-05-21 · arXiv · Credibility S

Leveraging continuous solar energy

Abstract, interpretation and reference

Abstract

Leveraging continuous solar energy

中文解读

背景:AI 数据中心负载、功率密度和能源约束同步上升,液冷、热管理和数据中心能效正在成为智算中心设计的关键变量。问题:论文聚焦现有方案在效率、可靠性或工程协同上的瓶颈。方法:摘要显示作者采用文献摘要中的模型、实验或案例分析,把运行负载、冷却/能源系统和基础设施约束放在同一分析框架中。结果:研究重点指向冷却效率、能源利用或运维策略的改进方向。意义:对日报读者而言,它可用于判断液冷方案、热管理路线和高密度部署节奏。仍需结合全文实验条件、样本范围和成本假设核验。

参考文献

Zhanwei Wang, Huiling Yang, Min Sheng, 等. SpaceMoE: Realizing Distributed Mixture-of-Experts Inference over Space Networks[J/OL]. (2026-05-21)[2026-06-06]. https://arxiv.org/abs/2605.00515.

Full text
热管理与液冷 论文图示