<aside> <img src="/icons/home_lightgray.svg" alt="/icons/home_lightgray.svg" width="40px" /> Home

Lessons

Assignments

Schedule

Library

FAQs

Course Info

Map

</aside>

Course Description

This course is designed to provide a comprehensive introduction to datacenter networks and systems. Students will learn how to design data center infrastructure for big data analytics, cloud services (e.g., Amazon EC2, Microsoft Windows Azure, and Google App Engine), and distributed machine learning (e.g., AllReduce, in-network computing, and NVidia NCCL).

The goal of this course is to study the key technologies and new challenges in data center networking and systems. The course will include paper presentations, discussions, and projects. The papers will be selected from top networking and systems conferences, organized in a bottom-up manner, to cover network infrastructure, routing and load balancing, congestion control, flow scheduling, networked systems, and applications.

Prerequisite (optional): Introduction to Computer Networks

Materials

All course readings, assignments, and other important information will be available on the course website. Please make sure to check it regularly.

Assessments

Paper review

All students are expected to read the papers selected for discussion before class and to submit written reviews for five papers during the semester. Reviews should be emailed to prior to the specified deadlines.

Each review should include the following components:

1. Paper summary: Briefly summarize the paper. What are the main problems addressed? Do you consider these problems important? Comment on the novelty and technical depth of the work.
2. Strengths: List 2–4 bullet points highlighting the strengths of the paper. Each point can be explained in more detail in the Comments section.
3. Weaknesses: List 2–4 bullet points identifying the weaknesses or limitations of the paper. Each point should be elaborated on in the Comments section.
4. Detailed comments: Provide detailed and constructive comments. (guidelines: Keshav's paper)

Project proposal

Each student is required to propose a research plan related to datacenter networks and systems. The topic may be inspired by, but is not limited to, the papers discussed in this course. The objective of this project is to practice identifying research problems and developing novel ideas to address them.

The project proposal should be concise yet clear, and should include:

1. The background and motivation of the problem
2. The research problem and goals
3. A preliminary system design
4. An evaluation plan

Presentation

Two teams of students will be chosen to lead the presentation and discussion of the paper selected each week. One team will be designated the offense and the other the defense. (How to give a presentation? Randy Katz's guidelines.)

Each week, two teams of students will be selected to lead the presentation and discussion of the assigned paper. One team will be designated as the defense team and the other as the offense team. (Presentation guidelines: Randy Katz's guidelines)

• Defense team
• Offense team
• Debate

Participation

All participants are expected to actively engage in the discussion by sharing what they learned from the presentation and by providing their own evaluation of the paper. After each presentation, you should submit a brief response that includes:

1. A summary of the key takeaways from the presentation.
2. Your overall judgment of the paper, based on the arguments presented by both the defense team and the offense team. You should explain the reasons supporting your judgment.

The participation survey may also include a short quiz to assess your understanding of the assigned pre-class readings.

<aside> 💡

Each survey is worth 2 points. Students only need to complete surveys for up to five presentations. Additional submissions will not earn extra credit.

</aside>

Built in Flotion.