Library of vectorized functions written in assembler.
Matrix operations are a highlight of the package, covering multiplication, inversion, LU decomposition, singular value decomposition, two-dimensional FFT, eigenvalues, and complex number math. OptiVec offers graphical representation of data in Cartesian coordinates, and the vectorized implementation in Assembler delivers speeds 2-3 times faster than compiled C/C++ source code of the same functionality. In many cases, this speed boost also comes with improved numerical accuracy.
This version of OptiVec is compatible with Borland C/C++ 4.0 or higher or Borland C++ Builder with Windows 95, 98, NT. All OptiVec functions have been optimized for Pentium/Pentium II but will run on hardware down to 486DX. Overall, OptiVec provides users with a comprehensive package of Assembler-optimized functions that can be incredibly useful for those who need to perform complex mathematical computations efficiently and quickly.
Version 3.3: N/A