maybe_make_cuda()

template<Rank rank, typename Scalar = SDef>

void syten::GenericDenseTensorImpl::GenericDenseTensor< rank, Scalar >::maybe_make_cuda ( )

inline

Conditionally transforms the stored dense tensor into a CUDA dense tensor if CUDA support is compiled in, at least one CUDA device is enabled and the tensor is larger than the CUDA threshold.

References syten::Cuda::cuda_compiled(), syten::Cuda::cuda_enabled(), syten::EnvVars::cuda_size_to_gpu, syten::GenericDenseTensorImpl::GenericDenseTensor< rank, Scalar >::make_cuda(), and syten::GenericDenseTensorImpl::GenericDenseTensor< rank, Scalar >::size().

Referenced by syten::TensorBlock< rank >::maybe_make_dense_cuda().

Here is the call graph for this function:

Here is the caller graph for this function: