Emerging Computing Paradigms

CSE 291

Instructor: Tajana Simunic Rosing
Email: tajana.teach@eng.ucsd.edu * please put CSE291 in the email subject line *
Lectures: 3:00-4:00pm Wednesdays
Office hours: 11-12pm Thursdays

Admin: James Martinez

Course Overview and Goals

Big data demands efficient processing. Todays traditional Von Neumann architectures are not able to meet these demands.  We need to explore alternative computational models which accelerate and replace today's computing tasks, enabling more dramatic scaling that will make it possible to run big data sets in real time going forward.  In this course, we cover diverse alternative computing paradigms such as approximate computing, deep Learning, hyperdimensional computing, processing in memory, quantum computing, stochastic computing, and DNA computing.  The goal of this course is to develop a comprehensive understanding of these technology advances and to set the stage for research in these exciting new areas.

Grading:
Students taking the class for S/U need to sign up to prepare either an overview, or a summary of the most recent research contributions on a topic listed below. Selected topic should be communicated to the instructor by the end of 2nd week of classes. Lecture is due in ppt form to instructor via email by 1pm the day before student is due to present it in class. In addition, the students need to provide written feedback to all other other presenters.

Students taking the class for a letter grade need to, in addition to requirements listed above, submit a 6pg, IEEE style, 2 column format research paper on a topic related to this class by week 10 and present a short overview of their work during week 10 lecture time. Research topics should be selected with the approval of the instructor by the 2nd week of classes.

Schedule

Week 1. Introduction

Week 2. Approximation techniques for general purpose processors (Daniel)

• Reading: A Survey of Techniques for Approximate Computing

Week 3. Circuit-level approximation techniques (Sahand)

• Reading: Approximate Computing: An Emerging Paradigm For Energy-Efficient Design

Week 4. Deep Learning (Yunhui)

• Reading: Deep Learning

Week 5. Hyperdimensional Computing (Mohsen, Yong)

• Reading: Hyperdimensional Computing: An Introduction to Computing in Distributed Representation with High-Dimensional Random Vectors

Week 6. Processing in memory (Joonseop, Sharadhi)

• Reading: In-Memory Big Data Management and Processing: A Survey

Week 7. Quantum computing (Samuel, Sanjay)

• Reading: A Gentle Introduction to Quantum Computing

Week 8. Stochastic computing (Saransh, Ricardo)

• Reading: Survey of Stochastic Computing

Week 9. DNA computing (Minxuan, Keith)

• Reading: Introduction to DNA Computing

Week 10. Final student presentations