Introduction
AI computation increasingly runs on specialized hardware and edge devices to deliver real-time performance at scale. This course explores the processors, accelerators, and distributed systems that enable efficient AI workloads. Participants will learn about GPUs, TPUs, FPGAs, and edge inference techniques. Real-world examples illustrate how AI is deployed in IoT, mobile, and embedded systems. By the end, learners will understand the hardware landscape powering modern AI.
Course Objectives
- Understand AI compute architectures
- Explore accelerators like GPUs and TPUs
- Learn optimization techniques for inference
- Study edge deployment strategies
- Build small edge-AI applications
Target Audience
- Hardware engineers
- ML/AI engineers
- IoT developers
- Embedded systems designers
- Students in computing and robotics
Course Outline
- 5 Sections
- 0 Lessons
- 5 Days
Expand all sectionsCollapse all sections
- Day 1: AI Hardware Overview• CPUs vs. GPUs
• Parallel processing
• Accelerator types
• Memory considerations
• Case studies0 - Day 2: GPU & TPU Architectures• CUDA basics
• Tensor cores
• TPU design
• Performance bottlenecks
• Hands-on: Optimize model training0 - Day 3: Inference Optimization• Quantization
• Pruning
• Model compression
• Batch vs. streaming inference
• Hands-on: Optimize a model0 - Day 4: Edge Computing• Edge vs. cloud
• IoT architectures
• On-device ML frameworks
• Energy constraints
• Hands-on: Deploy to edge0 - Day 5: Real-World Applications• Smart homes
• Autonomous drones
• Industrial monitoring
• Security systems
• Capstone demo0
Instructor
Lessons:
