USTC-SYS Reading Group

Intercepted from the original page.


The Systems Reading Group at University of Science and Technology of China provides an opportunity for students, researchers and faculty to discuss and keep abreast of current research in systems research, and to learn by example “how to do high-quality systems research”.


The meeting is held at 7:00 pm every Wednesday online (Tencent Meeting), and should not be longer than 1 hour.

Available Conference

Operating System: EuroSys 2020EuroSys 2021USENIX ATC ‘21OSDI ‘21

Storage System: FAST ‘21

Networking System: SIGCOMM 2021NSDI ‘20NSDI ‘21

Database System: SIGMOD 2020VLDB 2020

Architectures/Language: ASPLOS 2021

Cloud Computing: SoCC ‘20

Machine Learning and Systems: MLSYS 2021

2019 CCF Recommendation List: 中国计算机学会推荐国际学术会议和期刊目录-2019.pdf


Si Wu, Cheng Li organizes USTC-SYS Reading Group, advise students to select high-quality papers and how to present a paper, and gives comments on students’ presentation skills and leads the technical discussions.

Weiqiang Yu (Assistant, Fall 2021) is responsible for setting up the paper discussion leader for next week by the end of each week, reminding the person of selecting a high-quality paper to read, and announcing the paper selection in the wechat group by every Monday.

Discussion Format

The primary focus of the discussion will be understanding what the paper says. With that in mind, the discussion leads should focus on clearly identifying

  • what is the problem being solved?
  • what are the challenges? what makes the problem difficult?
  • what are the key insights/techniques/lessons used to address those challenges? (i.e., how does it work)

We will also discuss other points, but we will try to keep these points as secondary concerns as much as possible

  • does it work? (i.e., evaluate the evaluation)
  • why does the problem matter? i.e., why should we (not) read the paper?
  • other points that arise
    Reference:How to read a paper

Presentation/slide format

Consider structuring the presentation around the key discussion points. When guiding the discussion, the goal should be to guide the rest of the group through the discussion points above (roughly in order). We strongly suggest that the number of slides should not exceed 40.

Paper selection

Discussion leads should propose a paper and justify why we should discuss the paper.

The current suggestion are papers covering system topics and published at top system conferences such as OSDI, SOSP, ASPLOS, SIGMOD, VLDB, ATC, EuroSys, NSDI, FAST etc in the recent 3 years. Please send your choice to to get approval.

Program and Proceedings

Not included in this page.