2 papers
cs.PL2026
Numerical Superoptimization for Library Learning
Jonas Regehr, Mitch Briles, Zachary Tatlock +1
Numerical software depends on fast, accurate implementations of mathematical primitives like sin, exp, and log. Modern superoptimizers can optimize floating-point kernels against a…
cs.PL2024
Target-Aware Implementation of Real Expressions
Brett Saiki, Jackson Brough, Jonas Regehr +5
New low-precision accelerators, vector instruction sets, and library functions make maximizing accuracy and performance of numerical code increasingly challenging. Two lines of wor…