Home  /  Products  /  Stata 19  /  Basic Linear Algebra Subprograms (BLAS)

← See Stata 19's new features

Highlights

  • Improved Improved Basic Linear Algebra Subprograms (BLAS) implementation for macOS on Apple silicon using LAPACK/OpenBLAS

  • Settings to use the new optimized BLAS or revert to the previous implementation

Stata, when run on computers with Intel and AMD processors, uses the Intel Math Kernel Library (MKL), a highly optimized library of mathematical functions. MKL includes support for the Basic Linear Algebra Subprograms (BLAS), a library of routines for performing linear algebra operations such as vector and matrix multiplication. In fact, MKL provides a highly optimized implementation of BLAS that provides significant performance improvements. However, MKL is not available for Macs with Apple silicon. Stata for macOS on Apple silicon used an older implementation of BLAS that was not written to take full advantage of modern processor capabilities.

Stata for macOS on Apple silicon now uses LAPACK/OpenBLAS for numerical computations. OpenBLAS is an open-source implementation of BLAS and is optimized for modern hardware platforms, which makes it more current and faster than the previous LAPACK/BLAS implementation. OpenBLAS support is currently limited to Apple silicon, and the previous LAPACK/BLAS implementation is still available through a setting.


View all the new features in Stata 19.

Ready to get started?

Experience powerful statistical tools, reproducible workflows, and a seamless user experience—all in one trusted platform.