Love the idea of SIMD optimizations for special functions, but we can’t omit generic fallbacks. A big advantage of implementing these in Julia is automatic differentiation, and SIMD optimizations are a disadvantage for such purposes. Now, a SIMD-optimized dual numbers implementation…
Related topics
| Topic | Replies | Views | Activity | |
|---|---|---|---|---|
| [RFC] The future of Polynomials.jl | 1 | 920 | March 12, 2021 | |
| Any reason holds Julia from having a proper Pochhammer function? | 9 | 2412 | February 25, 2017 | |
| [ANN] SymbolicUtils.jl -- Groundwork for a symbolic ecosystem in Julia | 26 | 5201 | February 9, 2021 | |
| [ANN] SIMDMathFunctions.jl : Fast vectorized mathematical functions for SIMD.jl , using SLEEFPirates.jl | 10 | 1103 | March 18, 2024 | |
| Continued development of GSL.jl | 10 | 1610 | October 19, 2018 |