Benchmark-August2008

From Eigen
Revision as of 16:39, 17 March 2009 by Ggael (Talk | contribs)

(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search

This is an old benchmark provided for the record. See the lastest benchmark for further details.

Here is the list of the libraries that we compare:

  • eigen2: ourselves, with the default options (SSE2 vectorization enabled).
  • eigen2_novec: ourselves but with SSE2 vectorization entirely disabled.
  • INTEL_MKL: The Intel Math Kernel Library, which includes a BLAS/LAPACK. Closed-source.
  • GOTO: The GOTO BLAS library. Non-free license.
  • ATLAS: The math-atlas BLAS library
  • ublas: The Boost::uBLAS library.
  • mtl4: The Matrix Template Library v4.
  • blitz: The historical Blitz++ library, which introduced Expression Templates. The reason for its poor performance in this benchmark, is that it's simply not intended as a matrix library. We still include it here because many people misuse it in this way.
  • gmm: The GetFEM/GMM library.


Configuration

  • model name : Intel(R) Core(TM)2 CPU T7200 @ 2.00GHz ( x86_64 )
  • compiler: c++ (GCC) 4.2.3 (Ubuntu 4.2.3-2ubuntu7)
  • eigen2: Revision: 838022


axpy


axpby


matrix_vector


atv


matrix_matrix


aat


ata


trisolve


cholesky


hessenberg


tridiagonalization