AI Hardware Applications

This is the course website for ECE 495 - AI Hardware Applications, Fall ‘23.

Many of the pages are Jupyter Notebooks and can be opened in Google Colab by clicking the launch button 🚀 at the top right of the page!

Prediction

• 1. Hello, Colab!
• 2. Prediction Machines
• 3. K-Means clustering digits
• 4. Machine Learning Workflow
• 5. Dimensionality
• 6. C&C: Prediction and Dimensionality

Neural Networks

• 7. Neural Networks
• 8. Linear Algebra and GPUs
• 9. DNN Exploration
• 10. Convolutional Neural Network
• 11. Transfer Learning
• 12. TensorFlow Lite
• 13. TF Lite 🥧
• 14. Workday 1
• 15. Workday 2
• 16. Live Inference 🐱 🐶
• 17. C&C: Deep Neural Networks

NVIDIA Jetson

• 18. Containerization
• 19. Diffusion Image Generation
• 20. Containers Part 2
• 21. YOLO

Tensorflow Lite Micro

• 22. M4-Cortex Processor
• 23. Wake Word Detection

Natural Language Processing

• 24. Text Vectorization
• 25. Transformers
• 26. Emergence and Ethics of LLMs
• 27. Prompt Engineering

Final Block

• 28. Final Project
• 29. Final Project
• 30. Final Project
• 31. Final Project
• 32. Final Project
• 33. Super Computers
• 34. Final Project
• 35. Final Project
• 36. Bias and Ethics
• 37. Final C&C
• 38. LLM Red Team 1
• 39. LLM Red Team 2
• 40. Wrap and Critique

ECE 495: AI Hardware Applications 1.5 (1) This course will require students to engage with Artificial Intelligence algorithms and frameworks to train and deploy machine learning models. Students will explore which types of problems are appropriate for AI to solve, data pipelining, hardware accelerated training, and inference on hardware and power constrained devices. Assignments may include system design, programming, and book reviews. Classes are organized as lecture and laboratories devoted to applying models. Prereq: CompSci 110. Sem hrs: 1.5

Course Text

In addition to lots of online resources, we use TinyML by Pete Warden.

The book is available for free from O’Reilly via DoD MWR Libraries.