Course

CS6886W: Web M.Tech Systems Engineering for Deep Learning

A practical, project-based course on end-to-end systems design for Deep Learning, taught from the perspective of Computer Architecture.

Instructors
Gopalakrishnan Srinivasan SSB 202, Department of Computer Science and Engineering, IIT Madras
e-mail: sgopal@cse.iitm.ac.in
Introduction

This course is a practical and project based introduction to end-to-end systems design challenges in Deep Learning from the perspective of Computer Architecture. It covers topics from three fields: Machine Learning, Compilers, and Architecture. It targets self-motivated graduate students and upper-level undergraduates interested in pursuing the area in greater depth, potentially for a research project or thesis.

It is not a traditional lecture-and-exam course. Instead, it features in-class tutorials, hands-on assignments, and a significant course project. Students are expected to do original work that can potentially be extended even beyond the course itself.


"I hear and I forget. I see and I remember. I do and I understand."
- Confucius

Evaluation
Component Weight
3 assignments 15% (each)
Mid sem Exam (25 October 2026) 25%
End sem Exam (13 December 2026) 30%
Time Table & Location

Taught in the C slot

Day Time
Monday 16:00 – 17:15
Wednesday 16:00 – 17:15
Class Schedule (Tentative)
Week Dates Important Dates Topic and Lecture Slides Assignments Paper Reading
Week 1 Sep 7 - Sep 11 Introduction to Computer Architecture
Week 2 Sep 14 - Sep 18 Architecture cont.
Week 3 Sep 21 - Sep 25 Microprocessor-Evolution-OoO Multicore
Week 4 Sep 28 - Oct 2 Performance & Technology Scaling Amdahl's Law
Week 5 Oct 5 - Oct 10 Intro to Deep Learning ResNet paper
Week 6 Oct 12 - Oct 16 DNN Training Advances Assignment-1
Week 7 Oct 19 - Oct 23 DNN Performance & Pruning-Granularity
Week 8
Week 9
Week 10
Week 11
Week 12
Week 13
Week 14
Week 15
Week 16
Teaching Assistants
Rajshekhar Rakshit cs24s031@cse.iitm.ac.in