Rate this book

C++ Caching & Precomputation: Data Structure and Algorithm Optimizations

David Spuler

Rate this book

C++ Caching & Data Structure and Algorithm Optimizations is a book for C and C++ software engineers looking for advanced efficiency optimizations from the low-level hardware caches to building your own caches and precalculation optimizations at the algorithm and data structure level.

Highlights
- Precalculate an entire lookup table or perfect hash lookup.
- Modern C++ metaprogramming versions of search space exploration
- Source code pre-generation for offline compute in C code.

Table of Contents:Part Caching & Precomputation Introduction

1. Introduction to Caching & Precomputation

2. Theory of Caching & Precomputation

3. Lookup Table Precomputation

Part Compile-Time Precomputation

4. Metaprogramming Precomputation

5. Text File Embedding

6. Source Code Generation

7. Data Structure Unrolling

8. Unrolled Binary Search

9. Switch Multiway Branch Optimizations

10. Computed Goto Jump Tables

Part Algorithm Caching & Precomputation

11. Runtime Auxiliary Precomputation

12. Cache Locality

13. Cache Warming

14. Prefetching

15. LRU Cache Data Structure

16. Deferred Computation

17. Mathematical Algebraic Calculations

Part Precomputed Data Structures

18. Perfect Hashing

19. Old-Style Perfect Hashing in C

20. Prefilled Fixed-Size Data Structures

21. Loop Unrolling

22. Preallocated Memory Pools

23. Bloom Filters

Part General Compile-Time Techniques

24. Compile-Time Optimizations

25. Inline Functions

26. Recursion Flattening

Part General Efficiency Optimizations

27. Branchless Coding Techniques

28. Branchless Conditional Tests

29. Branchless Pitfalls

30. String Optimizations

Part LLM Caching & Precomputation Research

31. LLM Caching Research Overview

32. Semantic Caching

33. KV Caching Overview

34. Prefix KV Caching

35. Radix Trees

36. Non-Prefix KV Caching

37. FFN Fusion

38. FFN Folding

39. Vocabulary Trimming

40. LLM Kernel Synthesis

41. Weight Precomputation Research

Appendices

Appendix Long List of Low Latency Techniques

Appendix B. 500 LLM Inference Optimization Techniques

Appendix C. 200 CUDA C++ Optimization Techniques

579 pages, Kindle Edition

Published June 25, 2026