Logistics
Instructor: Christina Delimitrou (delimitrou@csail.mit.edu, Office: 32-G738)
Spring 2024, Tue/Thu 1:00-2:30pm, 5-134
Office hours: by appointment
Classroom: 5-134
Prerequisites: The class is open to PhD students as well as advanced undergraduate and MS students who have taken an advanced systems and/or architecture course (e.g., 6.191/6.004/6.033 or equivalent). If unsure about whether you meet the prerequisites, contact the instructor.
Textbook: The course is structured in a half-lectures/half-paper reading format. We will be using "The Datacenter as a Computer: An Introduction to the Design of Warehouse-Scale Machines, Third Edition", Luiz André Barroso, Urs Hölzle, Partha Ranganathan. Morgan & Claypool Publishers. Synthesis Lectures on Computer Architecture.
Projects: "Datacenter Computing" is a research-heavy course aimed at giving students early in their graduate (or late in their undergrad :)) career hands-on experience with real, cutting-edge research projects. As part of the course you will select a well-defined research problem, and work on it for the duration of the semester in groups of 2-3 students. For more information on the project as well as suggested research ideas see here.
Paper summaries: We will discuss up to two papers per class meeting. You are expected to read the papers ahead of the class meeting and submit a brief summary for each paper to delimitrou@csail.mit.edu. Consider the following questions in your summary: what is the problem addressed by the paper; what are the key insights they build upon; what are the major advantages or shortcomings of this work; what are the broader implications of this work (e.g., to higher/lower levels of the system stack, to key metrics, to similar problems,...).
Your summary should be roughly 1/2-1 page per paper. Submit a single PDF or TXT file with both summaries for the class meeting that is named <yourfamilyname>.<classtopic>.[pdf,txt]. Make sure you write you name in the file too.
Paper presentations: Each student will be assigned a paper to present during the class. You should prepare a 15-20 minute presentation on the paper, using slides (Powerpoint or Keynote). In your presentation, cover each of the following:
Problem: What is the paper trying to solve? How real is the problem?
Key idea: What is the main idea in the solution?
Novelty: What is different from previous work, and why? Is it a new problem, a new solution, or a new environment for an existing problem?
Critique: Is there anything you would change in the solution? What about in the way the authors presented or evaluated the solution?
Questions to the class: List a set of questions to start the class discussion on the paper. Questions can be things you wish the paper explained/tackled better, ideas on how you would expand on the work, or system implications the paper has in cloud computing overall.
Paper presentations can be found in this folder.
Grading: The main evaluation will be based on a research project that students propose and execute during the course. Apart from that, each student is expected to present one of the papers and to participate in class. The grading rubric will be 50% project, 25% midterm, 25% {summaries, paper presentation, and participation}.
You can find more information in the course syllabus.