[ Thread Index |
Date Index
| More lists.tuxfamily.org/eigen Archives
]
Christian wrote:
> @Jean Sreng: I don't know what your VR simulation is doing in detail.
> But for the normal 3D stuff you should be using small (3x3 or 4x4) fixed
> size matrices; those are perfectly supported by Eigen2 and using any
> BLAS implementation will only give you worse performance.
> If you are using big, variable sized matrices (e.g. to solve liear
> equation systems) the question for a BLAS backend is valid although it's
> worth a try to use the native implementation of Eigen2.
Yes, in fact we are using both tiny matrices (and native Eigen2 proved
to be efficient) and big matrices (much more than 1000) for specific
physics solvers. The main benefit we also see in using a BLAS
implementation inside of Eigen is to take benefit of either GPU
implementation (such as CUBLAS or similar libraries) or
multithread-optimized implementations (such as the Intel's MKL).
--
Jean