This chapter is divided into six parts and is meant to be read in order. Section numbers, theorem numbers, and exercise numbers are numbered uniformly across the whole chapter, so cross-part references remain valid as usual.
| Part | Article | Contents |
|---|---|---|
| Part One | Introduction, Specialization, Graph Capture, and Graph Transformations | Chapter summary; 1 Introduction; 2 Specialization and Caching; 3 Graph Capture; 4 Graph Transformations |
| Part Two | Automatic Differentiation and Lowering | 5 Automatic Differentiation and Activation Rematerialization; 6 Lowering and the Loop-Level Intermediate Representation |
| Part Three | Scheduling, Fusion, and CPU Code Generation | 7 Scheduling and Fusion; 8 CPU Code Generation |
| Part Four | GPU Code Generation and Dynamic Shapes | 9 GPU Code Generation; 10 Dynamic Shapes |
| Part Five | Caching, Parallel Compilation, Measurement, and Exercises | 11 Caching, Parallel Compilation, and Fallback; 12 Discussion; 13 Models and Measurement; Chapter Summary; Exercises; Programming Labs |
| Part Six | References and Appendices | References; Appendix A Source-Code Index; Appendix B Measurement Scripts; Appendix C Hints and Solutions for Selected Exercises |
