set_gpu_backend()

template<ComputeData InputType = std::vector<Kakshya::DataVariant>, ComputeData OutputType = InputType>

void MayaFlux::Yantra::ComputeOperation< InputType, OutputType >::set_gpu_backend ( std::shared_ptr< GpuExecutionContext< InputType, OutputType > >  backend )

inline

Attach a GPU execution backend.

When set and ready, apply_operation_internal delegates to the backend instead of operation_function. The CPU implementation in operation_function remains the automatic fallback when no backend is attached or GPU initialisation has not yet succeeded.

Parameters

backend

Configured GpuExecutionContext instance.

Definition at line 191 of file ComputeOperation.hpp.

    {
        m_gpu_backend = std::move(backend);
    }

References MayaFlux::Yantra::ComputeOperation< InputType, OutputType >::m_gpu_backend.

Referenced by MayaFlux::Yantra::GpuAnalyzer< InputType, OutputType >::GpuAnalyzer(), MayaFlux::Yantra::GpuExtractor< InputType, OutputType >::GpuExtractor(), MayaFlux::Yantra::GpuSorter< InputType, OutputType >::GpuSorter(), and MayaFlux::Yantra::GpuTransformer< InputType, OutputType >::GpuTransformer().

Here is the caller graph for this function: