Hello and welcome!

I am a tenure-track Assistant Professor in the Data Science and Analytics (DSA) Thrust at the Information Hub, The Hong Kong University of Science and Technology (Guangzhou), and an Affiliate Assistant Professor at The Hong Kong University of Science and Technology. I lead the Sustainable Computing (SusCom) Lab at HKUST(GZ), where our mission is to advance computing systems and AI Infra to enhance energy efficiency, scalability, and sustainability.

Current research interests include (Sustainable AI track):

  • KV cache optimization for memory-efficient LLM inference;
  • Energy-efficient AI model/LLM serving and deployment;
  • Scheduling, dispatching and orchestrating of computational resources (like GPUs) in cloud and edge environments.

We are also interested in developing AI-aided solutions for power monitoring, modeling and control in data centers and smart homes/buildings. Typical research problems that we are tackling include (AI for Sustainability track):

  • Software-defined power supplying for low/zero-carbon data centers;
  • Non-intrusive load monitoring (NILM) for smart homes;
  • Integrating LLM capability into building energy systems (BESs).
Recruitment: I am actively seeking self-motivated PhD students, RAs, and Interns to join my research group. If you are interested, please send your CV and transcripts to my email below. If you are currently (or going to be) enrolled in the RBM program at HKUST(GZ) and are interested in any of our research topics, please feel free to reach out 😊

Email: guomingtang [at] hkust-gz [dot] edu [dot] cn

News

  • May. 2025: LLM4NILM, the first study applying prompt-driven, general-purpose LLMs to NILM, is now available on arXiv.
  • May. 2025: Two papers got accepted to IEEE/ACM TON and IEEE TITS, respectively.
  • Apr. 2025: ZSMerge (zero-shot KV Cache Compression for LLMs) is now available on arXiv and open-sourced on GitHub.
  • Mar. 2025: GreenFL got accepted to ICDCS'25.
  • Mar. 2025: Three papers got accepted to IEEE TIFS, TGCN and TVT, respectively.
  • Nov. 2024: Our paper on real-world NILM system deployment won the Best Paper Award of SustainCom'24.
  • Jun. 2024: Our work on low-carbon edge computing system got accepted to ICPP'24.
    • (More...)