High Performance Linear Algebra in Scala

Please log in to watch this conference skillscast.

Netlib-java is for accessing hardware optimised linear algebra libraries (i.e. high performance mathematics). It is being used in investment banks, aerospace companies and has recently been adopted by the Apache Spark ecosystem:

https://github.com/fommil/netlib-java

In netlib-java, implementations of BLAS/LAPACK/ARPACK are provided by:

• delegating builds that use machine optimised system libraries (see below)
• self-contained native builds using the reference Fortran from netlib.org
• F2J to ensure full portability on the JVM

In this talk, Sam will be explaining why high performance linear algebra requires hardware support and show how to utilise netlib-java in Scala (e.g. via Breeze).

Sam will also discuss why GPUs are the future for linear algebra, and the technical challenges that remain.

YOU MAY ALSO LIKE:

• Workshop: Shapeless for Mortals (SkillsCast recorded in December 2015)
• Taming the Context Beast (SkillsCast recorded in October 2022)
• The Middle Way for Static Typing in Spark DataFrames (SkillsCast recorded in October 2022)