Production AI Systems
Learn the training, memory, inference, optimization, serving, and observability trade-offs behind production AI workloads.
Begin chapterPrerequisites
- 01Attention Variants & Norms
Grouped-Query Attention (GQA), Multi-Query (MQA), RoPE embeddings, and RMSNorm topology.
- 02Distributed Training
Parallelism paradigms (DDP, TP, PP, FSDP) and scaling laws.
- 03FlashAttention & GPU Architecture
GPU HBM vs SRAM bandwidth, online softmax tiling, IO-awareness, and Triton CUDA ops.
- 04Quantization & LoRA
Low-Rank Adaptation (ΔW = B·A), rank selection, FP8/INT4 uniform quantization, and scale-zero point math.
- 05Mixture of Experts (MoE)
Sparse routing mechanisms, gating networks, Top-K dispatch, and auxiliary load balancing.
- 06State Space Models & Mamba
Continuous-time SSMs, selective scan parameters, discretization, and linear-time sequence modeling.
- 07Model Serving
High-throughput LLM hosting, KV caching, vLLM, speculative decoding, and quantization.