Lightning Talk: Simple Compile Time Dynamic Programming
summary
Uses constexpr functions with plain arrays and loops to run bottom-up dynamic programming at compile time, building a DP cost matrix and a split-point matrix and returning the split table. Demonstrates matrix chain multiplication and a compile-time graph shortest path example. Benchmarking compiles a 12-matrix chain and shows the naive runtime at 172 ms versus the compile-time-optimized parenthesization at 9 ms, proving the technique produces correct, large speedups.
AI-written summary · about
related talks
50mC++ Costless Abstractions: the compiler view
14mFast Linear Programming through Transprecision Computing on Small and Sparse Data
01h 06m“C++ Metaprogramming: Journey from simple to insanity and back"
01h 01m“How bad is Meta-Programming still today?"
01h 01mExpressive Compile Time Parsers
01h 03mUnderstanding The constexpr 2-Step
01h 18m“Variable Templates and Compile-Time Computation with C++14"
15mSealing Pointer-Based Optimizations Behind Pure Functions