Syllabus

Last Updated: June 1st, 2021

Educational Objectives

15-849 is a graduate-level course at the intersection of computer architecture, systems, and hardware security for the datacenter. The course is composed of readings for each class and a semester-long research project. Upon successful completion of this course, students should be able to:
  • Understand the state-of-the-art hardware architectures and systems that power modern datacenters, with emphasis on the intersection of the two layers.
  • Identify the security implications of shared hardware resources and develop an understanding of existing attack vectors, defense mechanisms, and their limitations.
  • Interpret, criticize, and propose constructive improvements of research talks and papers, with a focus on the opportunities and constraints of co-designing software and hardware, as well as their security and privacy implications.
  • Develop and evaluate complex computer architecture and systems with a focus on robust experimental methodology and cross-layer explanation of design decisions.

These set of skills will be helpful in your research when targeting top tier venues in computer architecture and systems or in industry when building the next state-of-the art systems. The course will also help early and senior PhD students in building a solid understanding of fruitful research opportunities and directions.

Focus on understanding the abstractions and interfaces of each layer and create bridges across research areas.

Prerequisites

  • The course is appropriate for doctoral students in systems-related areas.
  • Master and undergraduate students may be admitted with permission from the instructor.

Tools and Text

  • We will mostly use C/C++ and computer architecture simulators, some projects might require Python, Rust, or Go. There will be not text, but some course notes and papers may be distributed.

Grading Scheme

  1. Lab Assignment (1): 15%
  2. Seminar: 35%
    • Paper reviews (2 per week) - 20%
    • Presentation lead - 10%
    • Participation/Discussion (2 papers per course) – 5%
  3. Research Project: 50%
    • Proposal - 10%
    • Midterm Checkpoint - 10%
    • Final Presentation - 15%
    • Final Report - 15%
Bonus: If you submit a paper to a top tier research venue as selected by the instructor and gets accepted you automatically get an A.

Reading Assignments & Reviews

For each class, there are two assigned readings and potentially a set of optional readings. Each student is expected to read both assignments. Each student is required to submit a one paragraph review for only one or both readings as indicated in the schedule. Each review must include the following information:

  • A summary of the main idea of the paper (Three sentences)
  • Opportunities and challenges of a cross-layer design (Three sentences)
  • Limitations (Two sentences)
  • Potential Future Work (Two sentences)

Grading Policies

  • Reviews: Paper reviews are due by midnight the day before class.
  • Presentations: Paper presentations are due by the presentation lead 24h before the day of class.
  • Late-work: Generally, late work is not accepted. Exceptions will be made in case of emergency, and I will coordinate with you as needed.

Academic Integrity & Collaboration

  • Discussion of research papers and ideas among students is strongly encouraged.
  • Reviews: Paper reviews must be done individually and on your own writing. Plagiarism will not be tolerated.
  • Lab Assignment will be conducted in teams of two students. Discussion among teams is allowed. Students may use any online code and material with clear citations. Each team must provide a report of their own writing and code.
  • Project: The research project will be conducted in teams of two or three students. Discussion among teams is encouraged. Each team must provide their own proposal, presentations, code, and report.
  • See CMU’s Academic Integrity Policy for additional information.

Student Wellness

  • As a student, you may face a range of challenges that may affect your academic performance and daily life. CMU offers Counseling and Psychological Services (CaPS) to all students. CaPS is confidential and treatment does work. Support is always available, get in touch at 412-268-2922. Additional resources can be found here.
  • The Computer Architecture Student Association (CASA) is an independent student-run organization with the express purpose of developing and fostering a positive and inviting student community within computer architecture. Created by students for students, CASA aims to support the student community throughout the demanding years of academic study.