odeint thrust slows down massively when using nested zip_iterators -

odeint thrust slows down massively when using nested zip_iterators -

i building analysis tool on top of odeint , thrust. tool solves big number of initial status problems. have had success next odeint , thrust tutorials / demos. until now, systems have been working have had less 5 state variables, need extend scheme capable of handling more.

as mentioned in tutorial, there little problem thrust's tuples can have 10 items. problem can solved using nested zip_iterators.

quoting near bottom of one of tutorials:

"a little difficulty thrust's tuples have maximal arity of 10. little problem since can create zip iterator packed zip iterators. top level zip iterator contains 1 zip iterator state, 1 normal iterator parameter, , 1 zip iterator derivative."

i have implemented solution, , works, unfortunately massive loss in speed. same scheme (a two-variable system, simultaneously solving 8192 initial conditions), original, simple not-extensible more 5 variables solution runs in on second:

real 0m1.244s user 0m0.798s sys 0m0.356s

whereas more complex, extensible nested solution takes ~2000 times longer!

real 4m3.613s user 2m15.124s sys 1m47.363s

the difference between 2 programs

inside functor's operator(), refer state variables , derivatives, and

inside functor's constructor, for_each command, create zip_iterators , tuples.

below have included excerpts these sections, each program. hope doing wrong here, because kind of speed loss devastating! help much appreciated.

excerpts "simple" code (non-nested-iterators) ////////////////////////////////////////////////////////// //// within icfunctor's operator() // getting state variable value_type x = thrust::get< 0 >( t ); // setting derivative thrust::get< 2 >( t ) = 0.5*a - 1.5*x; ////////////////////////////////////////////////////////// //// for_each statement creates zip_iterator thrust::for_each( //// start indices thrust::make_zip_iterator( thrust::make_tuple( // state-variables boost::begin( x ) + 0*m_n, boost::begin( x ) + 1*m_n, // derivatives boost::begin( dxdt ) + 0*m_n, boost::begin( dxdt ) + 1*m_n)) , //// end indices thrust::make_zip_iterator( thrust::make_tuple( // state-variables boost::begin( x ) + 1*m_n, boost::begin( x ) + 2*m_n, // derivatives boost::begin( dxdt ) + 1*m_n, boost::begin( dxdt ) + 2*m_n)) , icfunctor() ); excerpts "extensible" code (nested-iterators) ////////////////////////////////////////////////////////// //// within icfunctor's operator() // getting state variable const int state_variables = 0; // defined global constant value_type x = thrust::get<0>(thrust::get<state_variables>( t )); // setting derivative const int derivatives = 1; // defined global constant thrust::get<0>(thrust::get<derivatives>( t )) = 0.5*a - 1.5*x; ////////////////////////////////////////////////////////// //// for_each statement creates zip_iterator thrust::for_each( //// start indices thrust::make_zip_iterator( thrust::make_tuple( // state variables thrust::make_zip_iterator( thrust::make_tuple( boost::begin( x ) + 0*m_n, boost::begin( x ) + 1*m_n)), // derivatives thrust::make_zip_iterator( thrust::make_tuple( boost::begin( dxdt ) + 0*m_n, boost::begin( dxdt ) + 1*m_n)) )), //// end indices thrust::make_zip_iterator( thrust::make_tuple( // state variables thrust::make_zip_iterator( thrust::make_tuple( boost::begin( x ) + 1*m_n, boost::begin( x ) + 2*m_n)), // derivatives thrust::make_zip_iterator( thrust::make_tuple( boost::begin( dxdt ) + 1*m_n, boost::begin( dxdt ) + 2*m_n)) )), icfunctor() );

thrust odeint