Mark Liu

Mark Liu

PhD Student at Nanyang Technological University. Researching interesting things in Green Computing.

Xiyuan Liu

I am a Master by Research student in Computer Science and Engineering at Nanyang Technological University, Singapore, where I work with Yonggang Wen on sustainable computing and resilient communication systems. My current projects emphasize energy-aware infrastructure for large-scale data centers, including compute-power-thermal co-optimization, and the design of reliable network architectures that balance performance with environmental impact.

You can reach me at xiyuan005@e.ntu.edu.sg or on LinkedIn as MarkLiu.

Research Interests

  • Green Computing
  • Resilient Systems
  • Digital twin systems for large-scale computing infrastructure

Education

  • Nanyang Technological University, Singapore - Master by Research in Computer Science and Engineering, 2025-present
  • Xidian University, Shaanxi, China - B.S. in Telecommunication Engineering, 2021-2025

Selected Publications

  • X. Liu et al. “CountingStars: Low-overhead Network-wide Measurement in LEO Mega-constellation Networks,” arXiv:2508.13512, submitted to INFOCOM 2026.
  • X. Liu et al. “FauTE: Fault-tolerant Traffic Engineering in Data Center Network,” 9th Asia-Pacific Workshop on Networking (APNet), 2025.
  • L. Qin*, X. Liu* et al. “Satformer: Accurate and Robust Traffic Data Estimation for Satellite Networks,” NeurIPS 2024. (*equal contribution)
  • X. Liu et al. “Topo Ten: Topology-Driven Accurate Traffic Data Estimation for LEO Mega-Constellation Networks,” IEEE GLOBECOM 2024.

Research Experience

  • Fast and Lossless Link Switching for Opto-electronic Interconnects in AI Clusters (Dec 2024-Jun 2025) - Developed traffic-aware link reconfiguration strategies to alleviate congestion during large-scale distributed training.
  • Fault-tolerant Traffic Engineering in Data Center Networks (Sep 2024-May 2025) - Built deep learning based controllers that integrate link failure probabilities into routing decisions, improving throughput under failure scenarios.
  • CountingStars: Network-wide Measurement in LEO Mega-constellations (Jan 2025-Aug 2025) - Designed a digital twin driven measurement framework that reduces on-satellite memory usage and measurement error.
  • Satformer and Topo Ten for Satellite Traffic Estimation (Aug 2023-Aug 2024) - Co-developed adaptive sparse spatio-temporal attention mechanisms that capture nonlinear traffic patterns across dynamic satellite topologies.

Industry Projects

  • Digital Twin of Data Center (Jun 2025-present), OCBC - Leading the design of digital twin systems that cut data hall energy costs by an estimated 60,000 SGD annually.
  • Cloud Core Network Operations (Sep 2024-Nov 2024), Huawei - Evaluated 5G network slicing performance and led troubleshooting to stabilize deployments.
  • Heterogeneous Computing Expansion Platform (Dec 2022-May 2023), AMD Xilinx - Implemented routing, DMA, and ethernet modules on Zynq UltraScale+ for mesh-based inter-chip communication.
  • Fitness Manager (Dec 2022-Dec 2023), Hunan Chima Hongfan Co., Ltd. - Co-founded a computer vision based fitness assistant that attracted 13,000 daily users and secured 300,000 RMB in investment.

Honors and Awards

  • Bao Weiming Scholarship (6 of 6000 students), 2025
  • Baoyan Duan Science and Education Innovation Scholarship (4 of 6000 students), 2024
  • National Undergraduate Electronic Design Competition, 1st Prize (top 1% of 20,000+ teams), 2023
  • National College Students FPGA Innovative Design Competition, 2nd Prize, 2022
  • National College Students Embedded Chip Design and Application Competition, 2nd Prize, 2022
  • National College Students Mathematical Contest in Modeling, 2nd Prize, 2022
  • Xidian University Undergraduate Competition Winners Recognition Award, 2021-2023

Skills

  • Programming: C, C++, Python, Java, Golang, Verilog, MATLAB, Assembly
  • Domains: Computer networking, operating systems, computer architecture, optimization