Jihao Xin

Jihao Xin

Ph.D. Student in Machine Learning Systems

KAUST, Saudi Arabia

👋 I’m a second-year PhD student in computer science, working with Prof. Marco Canini at SANDS Lab. My research interests lie between computer systems and machine learning, devoting to speedup deep learning training and inference via efficient communication. My recent work focuses on gradient compression and optimized GPU communication.

Paper

Quickly discover relevant content by filtering papers.
(2023). Kimad: Adaptive Gradient Compression with Bandwidth Awareness. DistributedML'23.

PDF Cite DOI

(2023). Global QSGD: Practical Floatless Quantization for Distributed Learning with Theoretical Guarantees. Preprint (Poster accepted at Eurosys'23).

PDF Cite Poster DOI

(2022). A Convolutional Neural Network Based Maximum Power Point Voltage Forecasting Method for Pavement PV Array. TIM'22 (JCR Q1).

PDF Cite DOI

(2021). OpenFunction for Software Defined IoT. ISNCC'21 Best Paper.

PDF Cite DOI

Education

 
 
 
 
 
KAUST
Ph.D. student in Computer Science
January 2023 – Present Thuwal, Saudi Arabia
 
 
 
 
 
KAUST
Master of Science in Computer Science (GPA = 3.90/4.0, Dean’s List)
August 2021 – December 2022 Thuwal, Saudi Arabia
 
 
 
 
 
Imperial College London
Master of Science in Applied Computational Science and Engineering (Distinction)
October 2020 – October 2021 London, UK
 
 
 
 
 
Shandong University
Bachelor of Engineering in Computer Science and Technology (GPA = 89.34/100)
September 2016 – June 2020 Weihai, China

Internship

 
 
 
 
 
Microsoft Research Asia
Research Intern
Microsoft Research Asia
September 2023 – Present Beijing, China
 
 
 
 
 
ISTAustria
Research Intern
ISTAustria
May 2022 – August 2022 Klosterneuburg, Austria
TopK gradient compression with efficient threshold selection by CUDA/C++.
 
 
 
 
 
Tencent Cloud
Engineering Intern
Tencent Cloud
August 2019 – November 2019 Shenzhen, China
Tencent international website backend developer by GoLang.
 
 
 
 
 
The University of Western Australia
Research Intern
The University of Western Australia
June 2019 – August 2019 Remote
YOLO-based splash detection of Western Australia’s coastal highway by Python.

Experience

KAUST
TA - UESTC & KAUST Summer School
KAUST
TA - UESTC & KAUST Summer School
ICL
Robotics & AI Winter School - Imperial College London Hamlyn Center

I’m into …

Linux
C++/CUDA
Python/PyTorch

Gallery of KAUST

Contact

  • 966560569149
  • KAUST Campus, Thuwal, Mekka Province 23955
  • 4307-WS18, 4th Floor, Building 1