 |
CMSC 5743 Efficient Computing of Deep Neural Networks – Spring 2023
| Lecture: | H 18:30-21:15 | YIA LT9 |
| Instructor: | Prof. Bei Yu | byu@cse.cuhk.edu.hk |
| Course Tutors: | Wenqian Zhao | wqzhao@cse.cuhk.edu.hk |
| Wanli Chen | wlchen@cse.cuhk.edu.hk | |
| Yang Bai | ybai@cse.cuhk.edu.hk | |
| Xinyun Zhang | xyzhang21@cse.cuhk.edu.hk | |
| Xufeng Yao | xfyao@cse.cuhk.edu.hk |
Announcements
Course Requirements
Homework (30%), Lab (40%), quiz (30%), and extra bonus.
Please submit your lab reports and homework through blackboard (link).
Schedule
| Date | Topic | Lab/Tutorial/Quiz | Homework | Note |
| Jan. 12 | Introduction (slides) | Lab0 warm-up (slides) | ||
| Jan. 19 | — | — | Traveling | |
| Jan. 26 | — | — | Lunar New Year | |
| Feb. 02 | Im1 GEMM-1 (slides) | Lab1 GEMM-1 | HW 1 | |
| Feb. 09 | Mo1 Pruning (slides) | Tutorial-1 | ||
| Feb. 16 | Im2 GEMM-2 (slides) | Lab2 GEMM-2 | HW 2 | |
| Feb. 23 | Mo2 Decomposition (slides) | Tutorial-2 | ||
| Mar. 02 | Im3 GEMM-3 (slides) | — | HW 3 | |
| Mar. 09 | Mo3 Quantization (slides) | — | ||
| Mar. 16 | — | — | Lecturer in sick | |
| Mar. 23 | Mo4 BNN (slides) | Quiz-1, Lab3 GEMM-3 | ||
| Mar. 30 | Mo5 KD (slides) | |||
| Apr. 06 | Im4 CUDA-1 (slides) | |||
| Apr. 08 | Mo6 NAS (slides) | Tutorial-3 | HW 4 | |
| Apr. 13 | Im5 CUDA-2 | Lab4 CUDA-1 | ||
| Apr. 20 | Im6 TVM (slides) | Tutorial-4 | ||
| Apr. 27 | Quiz-2 | Lab5 TVM |
References
All papers mentioned in lecture slides.
Sze, Vivienne, et al., “Efficient processing of deep neural networks”, Synthesis Lectures on Computer Architecture, 2020.
Erwei Wang et al., “Deep Neural Network Approximation for Custom Hardware: Where We've Been, Where We're Going”, CSUR, 2019.
Qianru Zhang et al., “Recent Advances in Convolutional Neural Network Acceleration”, NeuroComputing, 2019.
Bing Li et al., “Running Sparse and Low-Precision Neural Network: When Algorithm Meets Hardware”, ASPDAC, 2018.
Important Dates:
11:59 AM, Feb. 09, 2023: HW 1 due.
11:59 PM, Feb. 15, 2023: Lab 1 due.
11:59 AM, Feb. 23, 2023: HW 2 due.
11:59 PM, Mar. 01, 2023: Lab 2 due.
11:59 PM, May 05, 2023: Lab 4 due.
Knowledge Graph
 |