prodD() [4/6]

template<Rank r, typename Scalar >

Scalar syten::CudaDenseTensorImpl::prodD ( CudaDenseTensor< r, Scalar > const &  a, DenseTensor< r, Scalar > const &  b, std::array< int, r > const &  c_a, std::array< int, r > const &  c_b, bool  conjugate = false  )

inline

Reordering scalar product of a CUDA dense tensor and a standard dense tensor.

If either tensor is larger than SYTEN_CUDA_THRESHOLD, the calculation is done on the GPU, otherwise, the calculation is done on the host.

References syten::EnvVars::cuda_size_to_gpu, and syten::CudaDenseTensorImpl::CudaDenseTensor< rank, Scalar >::size().

Here is the call graph for this function: