Correctly Rounded Functions For Vector Applications: A Performance Study
This work addresses the need for correctly rounded math functions in vector applications, ensuring cross-platform bitwise reproducibility, but the results are preliminary and the library is not yet available.
The authors designed SIMD algorithms for correctly rounded single-precision functions and integrated them into a CPU math library, aiming to create the first correctly rounded vector math library by mid-2026. They also adapted these implementations for GPUs and developed proof-of-concept SIMD implementations for two double-precision functions.
Following recent interest in correctly rounded math library functions (as currently recommended by the IEEE 754 standard), we have designed several SIMD algorithms for one-input single precision functions and integrated them into our CPU math library; these will form the core of the first correctly rounded vector math library, to be available to users in mid-2026. To take advantage of the cross-platform bitwise reproducibility afforded by correct rounding, we adapted and evaluated a few SIMD implementations on graphics processing units (GPU). In addition, we designed and evaluated proof-of-concept SIMD implementations of two correctly rounded double precision functions.