ICL Research Profile

ScaLAPACK

Overview

The Linear Algebra PACKage (LAPACK) and Scalable LAPACK (ScaLAPACK) are widely used libraries for solving dense linear algebra problems. ICL has contributed to the development and maintenance of these two packages. LAPACK is sequential, relies on the BLAS library, and benefits from the multi-core BLAS library. ScaLAPACK is parallel, distributed, and depends on the BLAS, LAPACK, MPI, and BLACS libraries.

The latest major release of LAPACK, version 3.11.0, was released in November 2022. This version added a number of improvements along with big fixes. Highlights include safe scaling algorithms for 2-norm and Givens rotation, multishift QZ algorithm with aggressive early deflation, Householder reconstruction, level-3 BLAS solver for triangular Sylvester equations, and faster algorithms for Least Squares. A new version of ScaLAPACK, version 2.2.0, was also released in 2022, and added ILP64 support.

Find out more at http://www.netlib.org/scalapack/

In Collaboration With

  1. University of California Berkeley
  2. University of Colorado Denver

Sponsored by

  1. Defense Advanced Research Projects Agency
  2. National Science Foundation
  3. The United States Department of Energy

Project Handout