Course Description

The future computing systems will spread from server to desktop, mobile to embedded, swarm of sensors to various wearables. The computing in these systems will be dominated by exponentially growing data from ubiquitous network (e.g., input from sensors), social media (e.g., images, videos), and scientific platforms (e.g., data generated by particle accelerators, DNA sequencing). Unfortunately, today’s systems are mainly bottlenecked by data storage and movement. Technology scaling and innovation in compute led to an imbalanced system where data movement is much more expensive than a unit computation (in terms of bandwidth, energy, and latency). This trend worsens with the difficulty in DRAM scaling that impacts capacity, latency, reliability, and cost of future memory systems.
It is time we rethink and redesign our computing model focusing on minimizing data movement and storage. The course will cover material from VLSI circuits, architecture design, programming languages, and new application domains. In this course, we will focus on process-in-memory architectures, heterogeneous and fine-grain integration of memory and computation, approximate computation, integration of persistent main memory systems, novel memory and compute technologies, and system designs driven by new applications, such as internet of things (IoT), virtual reality (VR), image processing, and bioinformatics.

Course Goals

Goal 1:

Critical analysis and in-class presentation and discussion of research papers in computer architecture: Students will be expected to read and present recent as well as established research papers, critically review and analyze them in writing, as well as discuss them in class. Most of the class time will be dedicated to such presentations and discussions. We expect at least two presentations per week.

Goal 2:

Semester-long research project: Students will be expected to propose and carry out a clearly-defined research project in computer architecture. The project is open-ended and has to be approved by the instructor. Project deliverables will consist of a project proposal writeup, milestone presentations and a final project report. Students are expected to conduct research projects that are potentially publishable.

Grade Distribution

Class Participation 10%
Reviews 25%
Class Presentations 30%
Project 35%

Meeting Times

Tuesday and Thursday 2.00 pm - 3.15 pm, Rice Hall 340

Contact

Instructor
Email: samira khan@virginia.edu
Office: Rice 308
Office Hour: By Appointment