You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Some ideas about a new architecture of Siconos Algebra.
The first tests with Petsc/TAp are really conclusive. We could think about a deep refactoring of Siconos Algebra. I can see an architecture like that:
a very lightweight C++ library of vectors and matrices (with namespace Siconos::Algebra::matrix) based on
on the development of Numerics (double * and NumericsMatrix) with only the dense and sparse types, on top of BLAS/LAPACK and csparse.
or on a completely separate from numerics in Eigen, tinyMatrix or blitz++.
In this way, we can the control of the memory management (pool and alloc, numpy wrap) for the modeling part
a serious HPC framework based on direct use of Petsc/TAO instead of the lightweight library in places where the problem is vectorized
reacted with thumbs up emoji reacted with thumbs down emoji reacted with laugh emoji reacted with hooray emoji reacted with confused emoji reacted with heart emoji reacted with rocket emoji reacted with eyes emoji
-
Hello,
Some ideas about a new architecture of Siconos Algebra.
The first tests with Petsc/TAp are really conclusive. We could think about a deep refactoring of Siconos Algebra. I can see an architecture like that:
Siconos::Algebra::matrix
) based onIn this way, we can the control of the memory management (pool and alloc, numpy wrap) for the modeling part
What do you think ?
Beta Was this translation helpful? Give feedback.
All reactions