Course Syllabus
15 weeks • 2-4 hours per week • Self-paced learning
Week 1: Energy & Data Center Infrastructure
Topic: The physical foundation of AI – power, cooling, and data center design
Time Estimate: 3-4 hours
Learning Goals
- Understand PUE (Power Usage Effectiveness) and data center efficiency metrics
- Learn about liquid cooling vs. air cooling for high-density compute
- Grasp the energy requirements for training large models
Week 2: Silicon Architecture – GPUs, TPUs, and Custom Chips
Topic: Understanding the hardware that powers AI at scale
Time Estimate: 3-4 hours
Learning Goals
- Compare GPU vs TPU vs custom ASIC architectures
- Understand tensor cores and specialized AI accelerators
- Learn about memory bandwidth bottlenecks and optimization
Week 3: Distributed Systems Fundamentals
Topic: The software infrastructure that coordinates thousands of GPUs
Time Estimate: 3-4 hours
Learning Goals
- Understand data parallelism vs model parallelism vs pipeline parallelism
- Learn about cluster management systems (Kubernetes, Slurm)
- Grasp networking requirements for distributed training
Week 4: Training Infrastructure at Scale
Topic: How to train models with thousands of GPUs across multiple data centers
Time Estimate: 3-4 hours
Learning Goals
- Learn about gradient synchronization strategies
- Understand mixed precision training and loss scaling
- Grasp checkpointing and fault tolerance in distributed training
Week 5: Neural Network Foundations
Topic: Building blocks of modern deep learning
Time Estimate: 4-5 hours
Learning Goals
- Master backpropagation and gradient descent variants
- Understand convolutional networks and their applications
- Learn about activation functions, normalization, and regularization
Week 6: The Transformer Revolution
Topic: Understanding the architecture that powers GPT, BERT, and modern LLMs
Time Estimate: 4-5 hours
Learning Goals
- Master self-attention mechanisms
- Understand positional encoding and multi-head attention
- Learn encoder vs decoder architectures
Week 7: Foundation Models – Architecture & Scaling
Topic: How GPT, Claude, Gemini, and other LLMs are built
Time Estimate: 4-5 hours
Learning Goals
- Understand GPT vs BERT vs T5 architectures
- Learn about scaling laws and emergent abilities
- Grasp the relationship between compute, data, and model size
Week 8: Training Foundation Models
Topic: The practical reality of training billion-parameter models
Time Estimate: 3-4 hours
Learning Goals
- Learn about data curation and preprocessing at scale
- Understand tokenization strategies
- Master learning rate schedules and warmup strategies
Week 9: Fine-tuning and Alignment
Topic: RLHF, instruction tuning, and making models useful
Time Estimate: 3-4 hours
Learning Goals
- Understand supervised fine-tuning (SFT) vs RLHF
- Learn about reward modeling and preference learning
- Grasp constitutional AI and safety alignment techniques
Week 10: Deployment and Inference Optimization
Topic: Serving models in production at scale
Time Estimate: 3-4 hours
Learning Goals
- Understand KV caching and PagedAttention
- Learn about quantization (INT8, INT4, GPTQ, AWQ)
- Master batching strategies and request scheduling
Week 11: AI Applications – From Models to Products
Topic: Building real-world applications on top of foundation models
Time Estimate: 3-4 hours
Learning Goals
- Learn about RAG (Retrieval Augmented Generation)
- Understand function calling and tool use
- Grasp prompt engineering and chain-of-thought reasoning
Week 12: Scientific AI Applications
Topic: AI transforming biology, chemistry, and scientific discovery
Time Estimate: 3-4 hours
Learning Goals
- Understand protein folding and AlphaFold
- Learn about drug discovery and molecular generation
- Grasp AI for climate modeling and materials science
Week 13: Business Models & Economics of AI
Topic: How AI companies make money and the investment landscape
Time Estimate: 3-4 hours
Learning Goals
- Understand API-based business models vs open source
- Learn about AI infrastructure costs and unit economics
- Grasp venture capital perspectives on AI investing
Week 14: Future Frontiers
Topic: What's next in AI infrastructure and applications
Time Estimate: 3-4 hours
Learning Goals
- Understand multimodal models (vision + language)
- Learn about agent-based AI systems
- Grasp the trajectory toward AGI and superintelligence
Week 15: Capstone Project
Topic: Apply everything you've learned to build your own frontier system
Time Estimate: 10-20 hours
Project Options
- Build a RAG chatbot using your own documents
- Fine-tune a small language model on a domain-specific corpus
- Implement a multi-agent AI system with tool use
- Deploy an optimized inference system with quantization
- Create a multimodal application combining vision and language