Software

High performance dense linear algebra libraries, each addressing a layer in the linear algebra software stack, have been developed by the team and our collaborators from both academia and industry.

BLIS is a software framework for instantiating high-performance BLAS-like dense linear algebra libraries. BLIS is written in Standard C (mostly ISO C90 with a few C99 extensions) and available under a new/modified/3-clause BSD license.
Get it here
libFLAME is a high performance dense linaer algebra library that is the result of the FLAME methodology for systematically developing dense linear algebra libraries. The FLAME methodology is radically different from the LINPACK/LAPACK approach that dates back to the 1970s.
libFLAME is LAPACK compatible and includes all of LAPACK's functionality.
Get it here
Elemental is a distributed-memory library for dense and sparse-direct linear algebra and optimization.
Get it here
ROTE is a library for tensor contractions targeting distributed memory architectures. ROTE is written in ... and available under a new/modified/3-clause BSD license.
Get it here
Get it here
DxTer is a domain-agnostic tool for generating high-performance code for an input specification using a knowledge base that encodes information about a particular software domain. DxTer has been used to generate code for BLIS, Elemental, and ROTE.
Get it here
The Linear Algebra Processor (LAP) cycle-accurate simulator models the bus transactions, state machine transitions, and computations of the LAP. The simulator engine is functional and performs the actual computations on the simulated hardware, making debugging under this environment easy.
Get it here