Welcome to Hardware/Software Co-Optimization for Machine Learning, taught by Prof. Luis Ceze with Thierry Moreau.

This website will be updated throughout the quarter, so check back for the latest.

Course Overview

Schedule

This is the planned lecture schedule. It will get refined as we go, so please check back.

Each week there will be required and suggested readings. Some of the class time will be use for paper presentations by students, paper discussions and hands-on tutorials.

(subject to change)

Date Topic & Readings HW
March 28 (W)

Intro, class planning

March 30 (F)

Intro to deep learning (MxNET tutorial)

April 4 (W)

Motivation for hardware specialization, Deep Learning accelerator taxonomy

April 6 (F)

DL accelerator optimizations, TPU deep dive

April 11 (W)

FPGA primer

April 13 (F)

HLS tutorial, tiny TPU (VTALite) lab overview

April 18 (W)

TVM intro (CPU+GPU)

April 20 (F)

VTA Design Optimizations

April 25 (W)

Supporting HW acceleration with TVM

April 27 (F)

Project Suggestions / Lab 2 Hackathon

May 2 (W)

Project pitches / Distributed training (Liang Luo)

May 4 (F)

Paper discussion (Joshua Fromm)

May 9 (W)

Paper discussion (Nancy Wang and Ollin Bohan)

May 11 (F)

Paper discussion (Sam Elliot and Meghan Cowan)

May 16 (W)

Project support session with Thierry

May 18 (F)

Dr. Eric Chung on Project Brainwave (Microsoft)

May 23 (W)

Paper discussion (James Bornholt and Julie Newcomb)

May 25 (F)

Paper discussion (Ali Najafi and Amrita Mazumdar)

May 30 (W)

Paper discussion (Mohamad Katanbaf and Vincent Lee)

June 1

Project presentations

Calendar

This calendar just has the lectures, office hours, and special activities/meetings to help you keep up. We'll always let you know about changes via email, but feel free to subscribe to this if you like.