| Course Name |
(中) AI晶片設計(3218) |
| (Eng.) Hardware Accelerators for AI Deep Neural Networks |
| Offering Dept |
Bachelor Program in Electrical Engineering and Computer Science |
| Course Type |
Elective |
Credits |
3 |
Teacher |
CHUNG-BIN WU |
| Department |
Bachelor Program in Electrical Engineering and Computer Science / Undergraduate |
Language |
Chinese |
英文/EMI |
|
Semester |
2025-FALL |
| Course Description |
本課程主要讓學生了解從AI模型設計至硬體架構實現上的相關技術與考量，包含DNN關鍵運算設計、DNN不同實現技術的評估、架構與資源的權衡、最佳化的作法、當前實作的挑戰與機會
This course mainly allows students to understand the relevant technologies and considerations from the design of AI models to the implementation of hardware architecture, including the key design considerations for DNNs, evaluation of different implementations technologies of DNN, tradeoffs between architectures and resources, the utility of various optimization approaches, and recent implementation trends and opportunities |
| Prerequisites |
|
self-directed learning in the course |
Y |
| Relevance of Course Objectives and Core Learning Outcomes(%) |
Teaching and Assessment Methods for Course Objectives |
| Course Objectives |
Competency Indicators |
Ratio(%) |
Teaching Methods |
Assessment Methods |
讓學生從AI模型的建立、架構的認識到設計實現有完整的學習歷程
To have a complete learning process from the establishment of AI models, recognition of architecture and design implementation. |
|
|
| Exercises |
| Other |
| Lecturing |
|
| Written Presentation |
| Attendance |
| Assignment |
| Quiz |
|
| Course Content and Homework/Schedule/Tests Schedule |
| Week |
Course Content |
| Week 1 |
AI 晶片設計簡介
Introduction to AI Accelerators |
| Week 2 |
深度神經網路概論
Overview of Deep Neural Networks |
| Week 3 |
DNN核心運算
DNN Kernel Computation |
| Week 4 |
DNN核心運算
DNN Kernel Computation |
| Week 5 |
DNN核心運算
DNN Kernel Computation |
| Week 6 |
DNN加速器
DNN Accelerators |
| Week 7 |
DNN加速器
DNN Accelerators |
| Week 8 |
期中考 |
| Week 9 |
DNN資料流設計
DNN Data Flow Design |
| Week 10 |
DNN資料流設計
DNN Data Flow Design |
| Week 11 |
DNN資料流設計
DNN Data Flow Design |
| Week 12 |
DNN模型與硬體共同設計
DNN Model and Hardware Co-Design |
| Week 13 |
DNN模型與硬體共同設計
DNN Model and Hardware Co-Design |
| Week 14 |
DNN模型與硬體共同設計
DNN Model and Hardware Co-Design |
| Week 15 |
DNN模型與硬體共同設計
DNN Model and Hardware Co-Design |
| Week 16 |
期末考
製作專題報告
製作專題報告 |
self-directed
learning |
|
|
| Evaluation |
| 期中考30%、作業/專題報告40%、期末考30% |
| Textbook & other References |
Efficient Processing of Deep Neural Networks (Morgan & Claypool)
Deep Learning - Hardware Design
http://eyeriss.mit.edu/tutorial.html
https://cs217.stanford.edu/ |
| Teaching Aids & Teacher's Website |
| iLeaning |
| Office Hours |
Wed. 11:00-17:00
Thur. 11:00-17:00 |
| Sustainable Development Goals, SDGs(Link URL) |
| 08.Decent Work and Economic Growth 09.Industry, Innovation and Infrastructure | include experience courses：N |
|