GitHub - achristensen07/hyperjama: an optimized linear algebra package improving on NIST's Template Numerical Toolkit

achristensen07 / hyperjama Public

Notifications You must be signed in to change notification settings
Fork 0
Star 2

an optimized linear algebra package improving on NIST's Template Numerical Toolkit

Notifications

Name		Name	Last commit message	Last commit date
Latest commit History 3 Commits
README		README
hyperjama_test.cpp		hyperjama_test.cpp
jama_cholesky.h		jama_cholesky.h
jama_lu.h		jama_lu.h
jama_qr.h		jama_qr.h
tnt_array2d_utils.h		tnt_array2d_utils.h
tnt_math_utils.h		tnt_math_utils.h

Repository files navigation

HyperJAMA

Α λ ε ξ

HyperJAMA is an open source high performance linear algebra extension to NIST's
Template Numerical Toolkit.  The same interface is used, but many of the functions
have been optimized using parallelization, vectorization, and cache alignment to
perform near the theoretical computational limits of modern processors.
BLAS functions are optimized using SSE and SSE2 operations.

Features
	Speed increases up to 128x
	Block parallel algorithms using OpenMP
	Better cache alignment
	Vectorized BLAS Level 1 Functions using SSE and SSE2 intrinsic functions
	Support for complex matrix types, such as TNT::Array2D<std::complex<double> >

Instructions
	1. Go to http://math.nist.gov/tnt/download.html and download the latest source
	   code of the Template Numerical Toolkit and JAMA/C++ Linear Algebra Package.
	2. Click here to download the HyperJAMA replacement headers.
	3. Replace the corresponding headers from your project with the HyperJAMA headers.
	4. Add OpenMP and SSE2 flags to compiler command line arguments. Instructions for
	   g++ and Visual Studio are printed to the console while compiling.