path: root/Source/Particles/Deposition/CurrentDeposition.H

/* Copyright 2019 Axel Huebl, David Grote, Maxence Thevenet
 * Remi Lehe, Weiqun Zhang, Michael Rowan
 *
 * This file is part of WarpX.
 *
 * License: BSD-3-Clause-LBNL
 */
#ifndef CURRENTDEPOSITION_H_
#define CURRENTDEPOSITION_H_

#include "Particles/Deposition/SharedDepositionUtils.H"
#include "ablastr/parallelization/KernelTimer.H"
#include "Particles/Pusher/GetAndSetPosition.H"
#include "Particles/ShapeFactors.H"
#include "Utils/TextMsg.H"
#include "Utils/WarpXAlgorithmSelection.H"
#include "Utils/WarpXConst.H"
#ifdef WARPX_DIM_RZ
#   include "Utils/WarpX_Complex.H"
#endif

#include "WarpX.H" // todo: remove include and pass globals as args

#include <AMReX.H>
#include <AMReX_Arena.H>
#include <AMReX_Array4.H>
#include <AMReX_REAL.H>

/**
 * \brief Current Deposition for thread thread_num
 * \tparam depos_order deposition order
 * \param GetPosition  A functor for returning the particle position.
 * \param wp           Pointer to array of particle weights.
 * \param uxp,uyp,uzp  Pointer to arrays of particle momentum.
 * \param ion_lev      Pointer to array of particle ionization level. This is
                         required to have the charge of each macroparticle
                         since q is a scalar. For non-ionizable species,
                         ion_lev is a null pointer.
 * \param jx_fab,jy_fab,jz_fab FArrayBox of current density, either full array or tile.
 * \param np_to_depose Number of particles for which current is deposited.
 * \param relative_time Time at which to deposit J, relative to the time of the
 *                      current positions of the particles. When different than 0,
 *                      the particle position will be temporarily modified to match
 *                      the time of the deposition.
 * \param dx           3D cell size
 * \param xyzmin       Physical lower bounds of domain.
 * \param lo           Index lower bounds of domain.
 * \param q            species charge.
 * \param n_rz_azimuthal_modes Number of azimuthal modes when using RZ geometry.
 * \param cost  Pointer to (load balancing) cost corresponding to box where present particles deposit current.
 * \param load_balance_costs_update_algo Selected method for updating load balance costs.
 */
template <int depos_order>
void doDepositionShapeN (const GetParticlePosition& GetPosition,
                         const amrex::ParticleReal * const wp,
                         const amrex::ParticleReal * const uxp,
                         const amrex::ParticleReal * const uyp,
                         const amrex::ParticleReal * const uzp,
                         const int* ion_lev,
                         amrex::FArrayBox& jx_fab,
                         amrex::FArrayBox& jy_fab,
                         amrex::FArrayBox& jz_fab,
                         long np_to_depose,
                         amrex::Real relative_time,
                         const std::array<amrex::Real,3>& dx,
                         const std::array<amrex::Real,3>& xyzmin,
                         amrex::Dim3 lo,
                         amrex::Real q,
                         int n_rz_azimuthal_modes,
                         amrex::Real* cost,
                         long load_balance_costs_update_algo)
{
    using namespace amrex::literals;

#if !defined(WARPX_DIM_RZ)
    amrex::ignore_unused(n_rz_azimuthal_modes);
#endif

#if !defined(AMREX_USE_GPU)
    amrex::ignore_unused(cost, load_balance_costs_update_algo);
#endif

    // Whether ion_lev is a null pointer (do_ionization=0) or a real pointer
    // (do_ionization=1)
    const bool do_ionization = ion_lev;
    const amrex::Real dzi = 1.0_rt/dx[2];
#if defined(WARPX_DIM_1D_Z)
    const amrex::Real invvol = dzi;
#endif
#if defined(WARPX_DIM_XZ) || defined(WARPX_DIM_RZ)
    const amrex::Real dxi = 1.0_rt/dx[0];
    const amrex::Real invvol = dxi*dzi;
#elif defined(WARPX_DIM_3D)
    const amrex::Real dxi = 1.0_rt/dx[0];
    const amrex::Real dyi = 1.0_rt/dx[1];
    const amrex::Real invvol = dxi*dyi*dzi;
#endif

#if (AMREX_SPACEDIM >= 2)
    const amrex::Real xmin = xyzmin[0];
#endif
#if defined(WARPX_DIM_3D)
    const amrex::Real ymin = xyzmin[1];
#endif
    const amrex::Real zmin = xyzmin[2];

    const amrex::Real clightsq = 1.0_rt/PhysConst::c/PhysConst::c;

    amrex::Array4<amrex::Real> const& jx_arr = jx_fab.array();
    amrex::Array4<amrex::Real> const& jy_arr = jy_fab.array();
    amrex::Array4<amrex::Real> const& jz_arr = jz_fab.array();
    amrex::IntVect const jx_type = jx_fab.box().type();
    amrex::IntVect const jy_type = jy_fab.box().type();
    amrex::IntVect const jz_type = jz_fab.box().type();

    constexpr int zdir = WARPX_ZINDEX;
    constexpr int NODE = amrex::IndexType::NODE;
    constexpr int CELL = amrex::IndexType::CELL;

    // Loop over particles and deposit into jx_fab, jy_fab and jz_fab
#if defined(WARPX_USE_GPUCLOCK)
    amrex::Real* cost_real = nullptr;
    if( load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock) {
        cost_real = (amrex::Real *) amrex::The_Managed_Arena()->alloc(sizeof(amrex::Real));
        *cost_real = 0._rt;
    }
#endif
    amrex::ParallelFor(
        np_to_depose,
        [=] AMREX_GPU_DEVICE (long ip) {
#if defined(WARPX_USE_GPUCLOCK)
            const auto KernelTimer = ablastr::parallelization::KernelTimer(
                cost && (load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock),
                cost_real);
#endif

            // --- Get particle quantities
            const amrex::Real gaminv = 1.0_rt/std::sqrt(1.0_rt + uxp[ip]*uxp[ip]*clightsq
                                                        + uyp[ip]*uyp[ip]*clightsq
                                                        + uzp[ip]*uzp[ip]*clightsq);
            amrex::Real wq  = q*wp[ip];
            if (do_ionization){
                wq *= ion_lev[ip];
            }

            amrex::ParticleReal xp, yp, zp;
            GetPosition(ip, xp, yp, zp);

            const amrex::Real vx  = uxp[ip]*gaminv;
            const amrex::Real vy  = uyp[ip]*gaminv;
            const amrex::Real vz  = uzp[ip]*gaminv;
            // wqx, wqy wqz are particle current in each direction
#if defined(WARPX_DIM_RZ)
            // In RZ, wqx is actually wqr, and wqy is wqtheta
            // Convert to cylinderical at the mid point
            const amrex::Real xpmid = xp + relative_time*vx;
            const amrex::Real ypmid = yp + relative_time*vy;
            const amrex::Real rpmid = std::sqrt(xpmid*xpmid + ypmid*ypmid);
            amrex::Real costheta;
            amrex::Real sintheta;
            if (rpmid > 0._rt) {
                costheta = xpmid/rpmid;
                sintheta = ypmid/rpmid;
            } else {
                costheta = 1._rt;
                sintheta = 0._rt;
            }
            const Complex xy0 = Complex{costheta, sintheta};
            const amrex::Real wqx = wq*invvol*(+vx*costheta + vy*sintheta);
            const amrex::Real wqy = wq*invvol*(-vx*sintheta + vy*costheta);
#else
            const amrex::Real wqx = wq*invvol*vx;
            const amrex::Real wqy = wq*invvol*vy;
#endif
            const amrex::Real wqz = wq*invvol*vz;

            // --- Compute shape factors
            Compute_shape_factor< depos_order > const compute_shape_factor;
#if (AMREX_SPACEDIM >= 2)
            // x direction
            // Get particle position after 1/2 push back in position
#if defined(WARPX_DIM_RZ)
            // Keep these double to avoid bug in single precision
            const double xmid = (rpmid - xmin)*dxi;
#else
            const double xmid = ((xp - xmin) + relative_time*vx)*dxi;
#endif
            // j_j[xyz] leftmost grid point in x that the particle touches for the centering of each current
            // sx_j[xyz] shape factor along x for the centering of each current
            // There are only two possible centerings, node or cell centered, so at most only two shape factor
            // arrays will be needed.
            // Keep these double to avoid bug in single precision
            double sx_node[depos_order + 1] = {0.};
            double sx_cell[depos_order + 1] = {0.};
            int j_node = 0;
            int j_cell = 0;
            if (jx_type[0] == NODE || jy_type[0] == NODE || jz_type[0] == NODE) {
                j_node = compute_shape_factor(sx_node, xmid);
            }
            if (jx_type[0] == CELL || jy_type[0] == CELL || jz_type[0] == CELL) {
                j_cell = compute_shape_factor(sx_cell, xmid - 0.5);
            }

            amrex::Real sx_jx[depos_order + 1] = {0._rt};
            amrex::Real sx_jy[depos_order + 1] = {0._rt};
            amrex::Real sx_jz[depos_order + 1] = {0._rt};
            for (int ix=0; ix<=depos_order; ix++)
            {
                sx_jx[ix] = ((jx_type[0] == NODE) ? amrex::Real(sx_node[ix]) : amrex::Real(sx_cell[ix]));
                sx_jy[ix] = ((jy_type[0] == NODE) ? amrex::Real(sx_node[ix]) : amrex::Real(sx_cell[ix]));
                sx_jz[ix] = ((jz_type[0] == NODE) ? amrex::Real(sx_node[ix]) : amrex::Real(sx_cell[ix]));
            }

            int const j_jx = ((jx_type[0] == NODE) ? j_node : j_cell);
            int const j_jy = ((jy_type[0] == NODE) ? j_node : j_cell);
            int const j_jz = ((jz_type[0] == NODE) ? j_node : j_cell);
#endif //AMREX_SPACEDIM >= 2

#if defined(WARPX_DIM_3D)
            // y direction
            // Keep these double to avoid bug in single precision
            const double ymid = ((yp - ymin) + relative_time*vy)*dyi;
            double sy_node[depos_order + 1] = {0.};
            double sy_cell[depos_order + 1] = {0.};
            int k_node = 0;
            int k_cell = 0;
            if (jx_type[1] == NODE || jy_type[1] == NODE || jz_type[1] == NODE) {
                k_node = compute_shape_factor(sy_node, ymid);
            }
            if (jx_type[1] == CELL || jy_type[1] == CELL || jz_type[1] == CELL) {
                k_cell = compute_shape_factor(sy_cell, ymid - 0.5);
            }
            amrex::Real sy_jx[depos_order + 1] = {0._rt};
            amrex::Real sy_jy[depos_order + 1] = {0._rt};
            amrex::Real sy_jz[depos_order + 1] = {0._rt};
            for (int iy=0; iy<=depos_order; iy++)
            {
                sy_jx[iy] = ((jx_type[1] == NODE) ? amrex::Real(sy_node[iy]) : amrex::Real(sy_cell[iy]));
                sy_jy[iy] = ((jy_type[1] == NODE) ? amrex::Real(sy_node[iy]) : amrex::Real(sy_cell[iy]));
                sy_jz[iy] = ((jz_type[1] == NODE) ? amrex::Real(sy_node[iy]) : amrex::Real(sy_cell[iy]));
            }
            int const k_jx = ((jx_type[1] == NODE) ? k_node : k_cell);
            int const k_jy = ((jy_type[1] == NODE) ? k_node : k_cell);
            int const k_jz = ((jz_type[1] == NODE) ? k_node : k_cell);
#endif

            // z direction
            // Keep these double to avoid bug in single precision
            const double zmid = ((zp - zmin) + relative_time*vz)*dzi;
            double sz_node[depos_order + 1] = {0.};
            double sz_cell[depos_order + 1] = {0.};
            int l_node = 0;
            int l_cell = 0;
            if (jx_type[zdir] == NODE || jy_type[zdir] == NODE || jz_type[zdir] == NODE) {
                l_node = compute_shape_factor(sz_node, zmid);
            }
            if (jx_type[zdir] == CELL || jy_type[zdir] == CELL || jz_type[zdir] == CELL) {
                l_cell = compute_shape_factor(sz_cell, zmid - 0.5);
            }
            amrex::Real sz_jx[depos_order + 1] = {0._rt};
            amrex::Real sz_jy[depos_order + 1] = {0._rt};
            amrex::Real sz_jz[depos_order + 1] = {0._rt};
            for (int iz=0; iz<=depos_order; iz++)
            {
                sz_jx[iz] = ((jx_type[zdir] == NODE) ? amrex::Real(sz_node[iz]) : amrex::Real(sz_cell[iz]));
                sz_jy[iz] = ((jy_type[zdir] == NODE) ? amrex::Real(sz_node[iz]) : amrex::Real(sz_cell[iz]));
                sz_jz[iz] = ((jz_type[zdir] == NODE) ? amrex::Real(sz_node[iz]) : amrex::Real(sz_cell[iz]));
            }
            int const l_jx = ((jx_type[zdir] == NODE) ? l_node : l_cell);
            int const l_jy = ((jy_type[zdir] == NODE) ? l_node : l_cell);
            int const l_jz = ((jz_type[zdir] == NODE) ? l_node : l_cell);

            // Deposit current into jx_arr, jy_arr and jz_arr
#if defined(WARPX_DIM_1D_Z)
            for (int iz=0; iz<=depos_order; iz++){
                amrex::Gpu::Atomic::AddNoRet(
                    &jx_arr(lo.x+l_jx+iz, 0, 0, 0),
                    sz_jx[iz]*wqx);
                amrex::Gpu::Atomic::AddNoRet(
                    &jy_arr(lo.x+l_jy+iz, 0, 0, 0),
                    sz_jy[iz]*wqy);
                amrex::Gpu::Atomic::AddNoRet(
                    &jz_arr(lo.x+l_jz+iz, 0, 0, 0),
                    sz_jz[iz]*wqz);
            }
#endif
#if defined(WARPX_DIM_XZ) || defined(WARPX_DIM_RZ)
            for (int iz=0; iz<=depos_order; iz++){
                for (int ix=0; ix<=depos_order; ix++){
                    amrex::Gpu::Atomic::AddNoRet(
                        &jx_arr(lo.x+j_jx+ix, lo.y+l_jx+iz, 0, 0),
                        sx_jx[ix]*sz_jx[iz]*wqx);
                    amrex::Gpu::Atomic::AddNoRet(
                        &jy_arr(lo.x+j_jy+ix, lo.y+l_jy+iz, 0, 0),
                        sx_jy[ix]*sz_jy[iz]*wqy);
                    amrex::Gpu::Atomic::AddNoRet(
                        &jz_arr(lo.x+j_jz+ix, lo.y+l_jz+iz, 0, 0),
                        sx_jz[ix]*sz_jz[iz]*wqz);
#if defined(WARPX_DIM_RZ)
                    Complex xy = xy0; // Note that xy is equal to e^{i m theta}
                    for (int imode=1 ; imode < n_rz_azimuthal_modes ; imode++) {
                        // The factor 2 on the weighting comes from the normalization of the modes
                        amrex::Gpu::Atomic::AddNoRet( &jx_arr(lo.x+j_jx+ix, lo.y+l_jx+iz, 0, 2*imode-1), 2._rt*sx_jx[ix]*sz_jx[iz]*wqx*xy.real());
                        amrex::Gpu::Atomic::AddNoRet( &jx_arr(lo.x+j_jx+ix, lo.y+l_jx+iz, 0, 2*imode  ), 2._rt*sx_jx[ix]*sz_jx[iz]*wqx*xy.imag());
                        amrex::Gpu::Atomic::AddNoRet( &jy_arr(lo.x+j_jy+ix, lo.y+l_jy+iz, 0, 2*imode-1), 2._rt*sx_jy[ix]*sz_jy[iz]*wqy*xy.real());
                        amrex::Gpu::Atomic::AddNoRet( &jy_arr(lo.x+j_jy+ix, lo.y+l_jy+iz, 0, 2*imode  ), 2._rt*sx_jy[ix]*sz_jy[iz]*wqy*xy.imag());
                        amrex::Gpu::Atomic::AddNoRet( &jz_arr(lo.x+j_jz+ix, lo.y+l_jz+iz, 0, 2*imode-1), 2._rt*sx_jz[ix]*sz_jz[iz]*wqz*xy.real());
                        amrex::Gpu::Atomic::AddNoRet( &jz_arr(lo.x+j_jz+ix, lo.y+l_jz+iz, 0, 2*imode  ), 2._rt*sx_jz[ix]*sz_jz[iz]*wqz*xy.imag());
                        xy = xy*xy0;
                    }
#endif
                }
            }
#elif defined(WARPX_DIM_3D)
            for (int iz=0; iz<=depos_order; iz++){
                for (int iy=0; iy<=depos_order; iy++){
                    for (int ix=0; ix<=depos_order; ix++){
                        amrex::Gpu::Atomic::AddNoRet(
                            &jx_arr(lo.x+j_jx+ix, lo.y+k_jx+iy, lo.z+l_jx+iz),
                            sx_jx[ix]*sy_jx[iy]*sz_jx[iz]*wqx);
                        amrex::Gpu::Atomic::AddNoRet(
                            &jy_arr(lo.x+j_jy+ix, lo.y+k_jy+iy, lo.z+l_jy+iz),
                            sx_jy[ix]*sy_jy[iy]*sz_jy[iz]*wqy);
                        amrex::Gpu::Atomic::AddNoRet(
                            &jz_arr(lo.x+j_jz+ix, lo.y+k_jz+iy, lo.z+l_jz+iz),
                            sx_jz[ix]*sy_jz[iy]*sz_jz[iz]*wqz);
                    }
                }
            }
#endif
        }
    );
#if defined(WARPX_USE_GPUCLOCK)
    if(cost && load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock) {
        amrex::Gpu::streamSynchronize();
        *cost += *cost_real;
        amrex::The_Managed_Arena()->free(cost_real);
    }
#endif
}

/**
 * \brief Current Deposition for thread thread_num using shared memory
 * \tparam depos_order deposition order
 * \param GetPosition  A functor for returning the particle position.
 * \param wp           Pointer to array of particle weights.
 * \param uxp,uyp,uzp  Pointer to arrays of particle momentum.
 * \param ion_lev      Pointer to array of particle ionization level. This is
                         required to have the charge of each macroparticle
                         since q is a scalar. For non-ionizable species,
                         ion_lev is a null pointer.
 * \param jx_fab,jy_fab,jz_fab FArrayBox of current density, either full array or tile.
 * \param np_to_depose Number of particles for which current is deposited.
 * \param dt           Time step for particle level
 * \param relative_time Time at which to deposit J, relative to the time of the
 *                      current positions of the particles. When different than 0,
 *                      the particle position will be temporarily modified to match
 *                      the time of the deposition.
 * \param dx           3D cell size
 * \param xyzmin       Physical lower bounds of domain.
 * \param lo           Index lower bounds of domain.
 * \param q            species charge.
 * \param n_rz_azimuthal_modes Number of azimuthal modes when using RZ geometry.
 * \param cost  Pointer to (load balancing) cost corresponding to box where present particles deposit current.
 * \param load_balance_costs_update_algo Selected method for updating load balance costs.
 */
template <int depos_order>
void doDepositionSharedShapeN (const GetParticlePosition& GetPosition,
                               const amrex::ParticleReal * const wp,
                               const amrex::ParticleReal * const uxp,
                               const amrex::ParticleReal * const uyp,
                               const amrex::ParticleReal * const uzp,
                               const int*  ion_lev,
                               amrex::FArrayBox& jx_fab,
                               amrex::FArrayBox& jy_fab,
                               amrex::FArrayBox& jz_fab,
                               long np_to_depose,
                               const amrex::Real relative_time,
                               const std::array<amrex::Real,3>& dx,
                               const std::array<amrex::Real,3>& xyzmin,
                               amrex::Dim3 lo,
                               amrex::Real q,
                               int n_rz_azimuthal_modes,
                               amrex::Real* cost,
                               long load_balance_costs_update_algo,
                               const amrex::DenseBins<WarpXParticleContainer::ParticleType>& a_bins,
                               const amrex::Box& box,
                               const amrex::Geometry& geom,
                               const amrex::IntVect& a_tbox_max_size)
{
    using namespace amrex::literals;

#if defined(AMREX_USE_HIP) || defined(AMREX_USE_CUDA)
    using namespace amrex;

    auto permutation = a_bins.permutationPtr();

    amrex::Array4<amrex::Real> const& jx_arr = jx_fab.array();
    amrex::Array4<amrex::Real> const& jy_arr = jy_fab.array();
    amrex::Array4<amrex::Real> const& jz_arr = jz_fab.array();
    amrex::IntVect const jx_type = jx_fab.box().type();
    amrex::IntVect const jy_type = jy_fab.box().type();
    amrex::IntVect const jz_type = jz_fab.box().type();

    constexpr int zdir = WARPX_ZINDEX;
    constexpr int NODE = amrex::IndexType::NODE;
    constexpr int CELL = amrex::IndexType::CELL;

    // Loop over particles and deposit into jx_fab, jy_fab and jz_fab
#if defined(WARPX_USE_GPUCLOCK)
    amrex::Real* cost_real = nullptr;
    if (cost && load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock) {
        cost_real = (amrex::Real *) amrex::The_Managed_Arena()->alloc(sizeof(amrex::Real));
        *cost_real = 0._rt;
    }
#endif
    const auto dxiarr = geom.InvCellSizeArray();
    const auto plo = geom.ProbLoArray();
    const auto domain = geom.Domain();

    amrex::Box sample_tbox(IntVect(AMREX_D_DECL(0,0,0)), a_tbox_max_size - 1);
    sample_tbox.grow(depos_order);

    amrex::Box sample_tbox_x = convert(sample_tbox, jx_type);
    amrex::Box sample_tbox_y = convert(sample_tbox, jy_type);
    amrex::Box sample_tbox_z = convert(sample_tbox, jz_type);

    const auto npts = amrex::max(sample_tbox_x.numPts(), sample_tbox_y.numPts(), sample_tbox_z.numPts());

    const int nblocks = a_bins.numBins();
    const int threads_per_block = WarpX::shared_mem_current_tpb;
    const auto offsets_ptr = a_bins.offsetsPtr();

    const std::size_t shared_mem_bytes = npts*sizeof(amrex::Real);
    const amrex::IntVect bin_size = WarpX::shared_tilesize;
    const std::size_t max_shared_mem_bytes = amrex::Gpu::Device::sharedMemPerBlock();
    WARPX_ALWAYS_ASSERT_WITH_MESSAGE(shared_mem_bytes <= max_shared_mem_bytes,
                                     "Tile size too big for GPU shared memory current deposition");

    amrex::ignore_unused(np_to_depose);
    // Launch one thread-block per bin
    amrex::launch(
            nblocks, threads_per_block, shared_mem_bytes, amrex::Gpu::gpuStream(),
            [=] AMREX_GPU_DEVICE () noexcept {
#if defined(WARPX_USE_GPUCLOCK)
        const auto KernelTimer = ablastr::parallelization::KernelTimer(
                cost && (load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock),
                cost_real);
#endif
        const int bin_id = blockIdx.x;
        const unsigned int bin_start = offsets_ptr[bin_id];
        const unsigned int bin_stop = offsets_ptr[bin_id+1];

        if (bin_start == bin_stop) { return; /*this bin has no particles*/ }

        // These boxes define the index space for the shared memory buffers
        amrex::Box buffer_box;
        {
            ParticleReal xp, yp, zp;
            GetPosition(permutation[bin_start], xp, yp, zp);
#if defined(WARPX_DIM_3D)
            IntVect iv = IntVect(int( amrex::Math::floor((xp-plo[0]) * dxiarr[0]) ),
                                 int( amrex::Math::floor((yp-plo[1]) * dxiarr[1]) ),
                                 int( amrex::Math::floor((zp-plo[2]) * dxiarr[2]) ));
#elif defined(WARPX_DIM_XZ) || defined(WARPX_DIM_RZ)
            IntVect iv = IntVect(int( amrex::Math::floor((xp-plo[0]) * dxiarr[0]) ),
                                 int( amrex::Math::floor((zp-plo[1]) * dxiarr[1]) ));
#elif defined(WARPX_DIM_1D_Z)
            IntVect iv = IntVect(int( amrex::Math::floor((zp-plo[0]) * dxiarr[0]) ));
#endif
            iv += domain.smallEnd();
            getTileIndex(iv, box, true, bin_size, buffer_box);
        }

        buffer_box.grow(depos_order);
        Box tbox_x = convert(buffer_box, jx_type);
        Box tbox_y = convert(buffer_box, jy_type);
        Box tbox_z = convert(buffer_box, jz_type);

        Gpu::SharedMemory<amrex::Real> gsm;
        amrex::Real* const shared = gsm.dataPtr();

        amrex::Array4<amrex::Real> const jx_buff(shared,
                amrex::begin(tbox_x), amrex::end(tbox_x), 1);
        amrex::Array4<amrex::Real> const jy_buff(shared,
                amrex::begin(tbox_y), amrex::end(tbox_y), 1);
        amrex::Array4<amrex::Real> const jz_buff(shared,
                amrex::begin(tbox_z), amrex::end(tbox_z), 1);

        // Zero-initialize the temporary array in shared memory
        volatile amrex::Real* vs = shared;
        for (int i = threadIdx.x; i < npts; i += blockDim.x){
            vs[i] = 0.0;
        }
        __syncthreads();
        for (unsigned int ip_orig = bin_start+threadIdx.x; ip_orig<bin_stop; ip_orig += blockDim.x)
        {
            const unsigned int ip = permutation[ip_orig];
            depositComponent<depos_order>(GetPosition, wp, uxp, uyp, uzp, ion_lev, jx_buff, jx_type,
                                          relative_time, dx, xyzmin, lo, q, n_rz_azimuthal_modes,
                                          ip, zdir, NODE, CELL, 0);
        }

        __syncthreads();
        addLocalToGlobal(tbox_x, jx_arr, jx_buff);
        for (int i = threadIdx.x; i < npts; i += blockDim.x){
            vs[i] = 0.0;
        }

        __syncthreads();
        for (unsigned int ip_orig = bin_start+threadIdx.x; ip_orig<bin_stop; ip_orig += blockDim.x)
        {
            const unsigned int ip = permutation[ip_orig];
            depositComponent<depos_order>(GetPosition, wp, uxp, uyp, uzp, ion_lev, jy_buff, jy_type,
                                          relative_time, dx, xyzmin, lo, q, n_rz_azimuthal_modes,
                                          ip, zdir, NODE, CELL, 1);
        }

        __syncthreads();
        addLocalToGlobal(tbox_y, jy_arr, jy_buff);
        for (int i = threadIdx.x; i < npts; i += blockDim.x){
            vs[i] = 0.0;
        }

        __syncthreads();
        for (unsigned int ip_orig = bin_start+threadIdx.x; ip_orig<bin_stop; ip_orig += blockDim.x)
        {
            const unsigned int ip = permutation[ip_orig];
            depositComponent<depos_order>(GetPosition, wp, uxp, uyp, uzp, ion_lev, jz_buff, jz_type,
                                          relative_time, dx, xyzmin, lo, q, n_rz_azimuthal_modes,
                                          ip, zdir, NODE, CELL, 2);
        }

        __syncthreads();
        addLocalToGlobal(tbox_z, jz_arr, jz_buff);
    });
#if defined(WARPX_USE_GPUCLOCK)
    if (cost && load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock) {
        amrex::Gpu::streamSynchronize();
        *cost += *cost_real;
        amrex::The_Managed_Arena()->free(cost_real);
    }
#endif
#else // not using hip/cuda
    // Note, you should never reach this part of the code. This funcion cannot be called unless
    // using HIP/CUDA, and those things are checked prior
    //don't use any args
    ignore_unused( GetPosition, wp, uxp, uyp, uzp, ion_lev, jx_fab, jy_fab, jz_fab, np_to_depose, relative_time, dx, xyzmin, lo, q, n_rz_azimuthal_modes, cost, load_balance_costs_update_algo, a_bins, box, geom, a_tbox_max_size);
    WARPX_ABORT_WITH_MESSAGE("Shared memory only implemented for HIP/CUDA");
#endif
}

/**
 * \brief Esirkepov Current Deposition for thread thread_num
 *
 * \tparam depos_order  deposition order
 * \param GetPosition  A functor for returning the particle position.
 * \param wp           Pointer to array of particle weights.
 * \param uxp,uyp,uzp  Pointer to arrays of particle momentum.
 * \param ion_lev      Pointer to array of particle ionization level. This is
                       required to have the charge of each macroparticle
                       since q is a scalar. For non-ionizable species,
                       ion_lev is a null pointer.
 * \param Jx_arr,Jy_arr,Jz_arr Array4 of current density, either full array or tile.
 * \param np_to_depose Number of particles for which current is deposited.
 * \param dt           Time step for particle level
 * \param[in] relative_time Time at which to deposit J, relative to the time of the
 *                          current positions of the particles. When different than 0,
 *                          the particle position will be temporarily modified to match
 *                          the time of the deposition.
 * \param dx           3D cell size
 * \param xyzmin       Physical lower bounds of domain.
 * \param lo           Index lower bounds of domain.
 * \param q            species charge.
 * \param n_rz_azimuthal_modes Number of azimuthal modes when using RZ geometry.
 * \param cost Pointer to (load balancing) cost corresponding to box where present particles deposit current.
 * \param load_balance_costs_update_algo Selected method for updating load balance costs.
 */
template <int depos_order>
void doEsirkepovDepositionShapeN (const GetParticlePosition& GetPosition,
                                  const amrex::ParticleReal * const wp,
                                  const amrex::ParticleReal * const uxp,
                                  const amrex::ParticleReal * const uyp,
                                  const amrex::ParticleReal * const uzp,
                                  const int* ion_lev,
                                  const amrex::Array4<amrex::Real>& Jx_arr,
                                  const amrex::Array4<amrex::Real>& Jy_arr,
                                  const amrex::Array4<amrex::Real>& Jz_arr,
                                  long np_to_depose,
                                  amrex::Real dt,
                                  amrex::Real relative_time,
                                  const std::array<amrex::Real,3>& dx,
                                  std::array<amrex::Real, 3> xyzmin,
                                  amrex::Dim3 lo,
                                  amrex::Real q,
                                  int n_rz_azimuthal_modes,
                                  amrex::Real * const cost,
                                  long load_balance_costs_update_algo)
{
    using namespace amrex;
    using namespace amrex::literals;

#if !defined(WARPX_DIM_RZ)
    ignore_unused(n_rz_azimuthal_modes);
#endif

#if !defined(AMREX_USE_GPU)
    amrex::ignore_unused(cost, load_balance_costs_update_algo);
#endif

    // Whether ion_lev is a null pointer (do_ionization=0) or a real pointer
    // (do_ionization=1)
    bool const do_ionization = ion_lev;
#if !defined(WARPX_DIM_1D_Z)
    Real const dxi = 1.0_rt / dx[0];
#endif
#if !defined(WARPX_DIM_1D_Z)
    Real const xmin = xyzmin[0];
#endif
#if defined(WARPX_DIM_3D)
    Real const dyi = 1.0_rt / dx[1];
    Real const ymin = xyzmin[1];
#endif
    Real const dzi = 1.0_rt / dx[2];
    Real const zmin = xyzmin[2];

#if defined(WARPX_DIM_3D)
    Real const invdtdx = 1.0_rt / (dt*dx[1]*dx[2]);
    Real const invdtdy = 1.0_rt / (dt*dx[0]*dx[2]);
    Real const invdtdz = 1.0_rt / (dt*dx[0]*dx[1]);
#elif defined(WARPX_DIM_XZ) || defined(WARPX_DIM_RZ)
    Real const invdtdx = 1.0_rt / (dt*dx[2]);
    Real const invdtdz = 1.0_rt / (dt*dx[0]);
    Real const invvol = 1.0_rt / (dx[0]*dx[2]);
#elif defined(WARPX_DIM_1D_Z)
    Real const invdtdz = 1.0_rt / (dt*dx[0]);
    Real const invvol = 1.0_rt / (dx[2]);
#endif

#if defined(WARPX_DIM_RZ)
    Complex const I = Complex{0._rt, 1._rt};
#endif

    Real const clightsq = 1.0_rt / ( PhysConst::c * PhysConst::c );
#if !defined(WARPX_DIM_1D_Z)
    Real constexpr one_third = 1.0_rt / 3.0_rt;
    Real constexpr one_sixth = 1.0_rt / 6.0_rt;
#endif

    // Loop over particles and deposit into Jx_arr, Jy_arr and Jz_arr
#if defined(WARPX_USE_GPUCLOCK)
    amrex::Real* cost_real = nullptr;
    if( load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock) {
        cost_real = (amrex::Real *) amrex::The_Managed_Arena()->alloc(sizeof(amrex::Real));
        *cost_real = 0._rt;
    }
#endif
    amrex::ParallelFor(
        np_to_depose,
        [=] AMREX_GPU_DEVICE (long const ip) {
#if defined(WARPX_USE_GPUCLOCK)
            const auto KernelTimer = ablastr::parallelization::KernelTimer(
                cost && (load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock),
                cost_real);
#endif

            // --- Get particle quantities
            Real const gaminv = 1.0_rt/std::sqrt(1.0_rt + uxp[ip]*uxp[ip]*clightsq
                                                 + uyp[ip]*uyp[ip]*clightsq
                                                 + uzp[ip]*uzp[ip]*clightsq);

            // wqx, wqy wqz are particle current in each direction
            Real wq = q*wp[ip];
            if (do_ionization){
                wq *= ion_lev[ip];
            }

            ParticleReal xp, yp, zp;
            GetPosition(ip, xp, yp, zp);

#if !defined(WARPX_DIM_1D_Z)
            Real const wqx = wq*invdtdx;
#endif
#if defined(WARPX_DIM_3D)
            Real const wqy = wq*invdtdy;
#endif
            Real const wqz = wq*invdtdz;

            // computes current and old position in grid units
#if defined(WARPX_DIM_RZ)
            Real const xp_new = xp + (relative_time + 0.5_rt*dt)*uxp[ip]*gaminv;
            Real const yp_new = yp + (relative_time + 0.5_rt*dt)*uyp[ip]*gaminv;
            Real const xp_mid = xp_new - 0.5_rt*dt*uxp[ip]*gaminv;
            Real const yp_mid = yp_new - 0.5_rt*dt*uyp[ip]*gaminv;
            Real const xp_old = xp_new - dt*uxp[ip]*gaminv;
            Real const yp_old = yp_new - dt*uyp[ip]*gaminv;
            Real const rp_new = std::sqrt(xp_new*xp_new + yp_new*yp_new);
            Real const rp_mid = std::sqrt(xp_mid*xp_mid + yp_mid*yp_mid);
            Real const rp_old = std::sqrt(xp_old*xp_old + yp_old*yp_old);
            Real costheta_new, sintheta_new;
            if (rp_new > 0._rt) {
                costheta_new = xp_new/rp_new;
                sintheta_new = yp_new/rp_new;
            } else {
                costheta_new = 1._rt;
                sintheta_new = 0._rt;
            }
            amrex::Real costheta_mid, sintheta_mid;
            if (rp_mid > 0._rt) {
                costheta_mid = xp_mid/rp_mid;
                sintheta_mid = yp_mid/rp_mid;
            } else {
                costheta_mid = 1._rt;
                sintheta_mid = 0._rt;
            }
            amrex::Real costheta_old, sintheta_old;
            if (rp_old > 0._rt) {
                costheta_old = xp_old/rp_old;
                sintheta_old = yp_old/rp_old;
            } else {
                costheta_old = 1._rt;
                sintheta_old = 0._rt;
            }
            const Complex xy_new0 = Complex{costheta_new, sintheta_new};
            const Complex xy_mid0 = Complex{costheta_mid, sintheta_mid};
            const Complex xy_old0 = Complex{costheta_old, sintheta_old};
            // Keep these double to avoid bug in single precision
            double const x_new = (rp_new - xmin)*dxi;
            double const x_old = (rp_old - xmin)*dxi;
#else
#if !defined(WARPX_DIM_1D_Z)
            // Keep these double to avoid bug in single precision
            double const x_new = (xp - xmin + (relative_time + 0.5_rt*dt)*uxp[ip]*gaminv)*dxi;
            double const x_old = x_new - dt*dxi*uxp[ip]*gaminv;
#endif
#endif
#if defined(WARPX_DIM_3D)
            // Keep these double to avoid bug in single precision
            double const y_new = (yp - ymin + (relative_time + 0.5_rt*dt)*uyp[ip]*gaminv)*dyi;
            double const y_old = y_new - dt*dyi*uyp[ip]*gaminv;
#endif
            // Keep these double to avoid bug in single precision
            double const z_new = (zp - zmin + (relative_time + 0.5_rt*dt)*uzp[ip]*gaminv)*dzi;
            double const z_old = z_new - dt*dzi*uzp[ip]*gaminv;

#if defined(WARPX_DIM_RZ)
            Real const vy = (-uxp[ip]*sintheta_mid + uyp[ip]*costheta_mid)*gaminv;
#elif defined(WARPX_DIM_XZ)
            Real const vy = uyp[ip]*gaminv;
#elif defined(WARPX_DIM_1D_Z)
            Real const vx = uxp[ip]*gaminv;
            Real const vy = uyp[ip]*gaminv;
#endif

            // Shape factor arrays
            // Note that there are extra values above and below
            // to possibly hold the factor for the old particle
            // which can be at a different grid location.
            // Keep these double to avoid bug in single precision
#if !defined(WARPX_DIM_1D_Z)
            double sx_new[depos_order + 3] = {0.};
            double sx_old[depos_order + 3] = {0.};
#endif
#if defined(WARPX_DIM_3D)
            // Keep these double to avoid bug in single precision
            double sy_new[depos_order + 3] = {0.};
            double sy_old[depos_order + 3] = {0.};
#endif
            // Keep these double to avoid bug in single precision
            double sz_new[depos_order + 3] = {0.};
            double sz_old[depos_order + 3] = {0.};

            // --- Compute shape factors
            // Compute shape factors for position as they are now and at old positions
            // [ijk]_new: leftmost grid point that the particle touches
            const Compute_shape_factor< depos_order > compute_shape_factor;
            const Compute_shifted_shape_factor< depos_order > compute_shifted_shape_factor;

#if !defined(WARPX_DIM_1D_Z)
            const int i_new = compute_shape_factor(sx_new+1, x_new);
            const int i_old = compute_shifted_shape_factor(sx_old, x_old, i_new);
#endif
#if defined(WARPX_DIM_3D)
            const int j_new = compute_shape_factor(sy_new+1, y_new);
            const int j_old = compute_shifted_shape_factor(sy_old, y_old, j_new);
#endif
            const int k_new = compute_shape_factor(sz_new+1, z_new);
            const int k_old = compute_shifted_shape_factor(sz_old, z_old, k_new);

            // computes min/max positions of current contributions
#if !defined(WARPX_DIM_1D_Z)
            int dil = 1, diu = 1;
            if (i_old < i_new) dil = 0;
            if (i_old > i_new) diu = 0;
#endif
#if defined(WARPX_DIM_3D)
            int djl = 1, dju = 1;
            if (j_old < j_new) djl = 0;
            if (j_old > j_new) dju = 0;
#endif
            int dkl = 1, dku = 1;
            if (k_old < k_new) dkl = 0;
            if (k_old > k_new) dku = 0;

#if defined(WARPX_DIM_3D)

            for (int k=dkl; k<=depos_order+2-dku; k++) {
                for (int j=djl; j<=depos_order+2-dju; j++) {
                    amrex::Real sdxi = 0._rt;
                    for (int i=dil; i<=depos_order+1-diu; i++) {
                        sdxi += wqx*(sx_old[i] - sx_new[i])*(
                            one_third*(sy_new[j]*sz_new[k] + sy_old[j]*sz_old[k])
                           +one_sixth*(sy_new[j]*sz_old[k] + sy_old[j]*sz_new[k]));
                        amrex::Gpu::Atomic::AddNoRet( &Jx_arr(lo.x+i_new-1+i, lo.y+j_new-1+j, lo.z+k_new-1+k), sdxi);
                    }
                }
            }
            for (int k=dkl; k<=depos_order+2-dku; k++) {
                for (int i=dil; i<=depos_order+2-diu; i++) {
                    amrex::Real sdyj = 0._rt;
                    for (int j=djl; j<=depos_order+1-dju; j++) {
                        sdyj += wqy*(sy_old[j] - sy_new[j])*(
                            one_third*(sx_new[i]*sz_new[k] + sx_old[i]*sz_old[k])
                           +one_sixth*(sx_new[i]*sz_old[k] + sx_old[i]*sz_new[k]));
                        amrex::Gpu::Atomic::AddNoRet( &Jy_arr(lo.x+i_new-1+i, lo.y+j_new-1+j, lo.z+k_new-1+k), sdyj);
                    }
                }
            }
            for (int j=djl; j<=depos_order+2-dju; j++) {
                for (int i=dil; i<=depos_order+2-diu; i++) {
                    amrex::Real sdzk = 0._rt;
                    for (int k=dkl; k<=depos_order+1-dku; k++) {
                        sdzk += wqz*(sz_old[k] - sz_new[k])*(
                            one_third*(sx_new[i]*sy_new[j] + sx_old[i]*sy_old[j])
                           +one_sixth*(sx_new[i]*sy_old[j] + sx_old[i]*sy_new[j]));
                        amrex::Gpu::Atomic::AddNoRet( &Jz_arr(lo.x+i_new-1+i, lo.y+j_new-1+j, lo.z+k_new-1+k), sdzk);
                    }
                }
            }

#elif defined(WARPX_DIM_XZ) || defined(WARPX_DIM_RZ)

            for (int k=dkl; k<=depos_order+2-dku; k++) {
                amrex::Real sdxi = 0._rt;
                for (int i=dil; i<=depos_order+1-diu; i++) {
                    sdxi += wqx*(sx_old[i] - sx_new[i])*0.5_rt*(sz_new[k] + sz_old[k]);
                    amrex::Gpu::Atomic::AddNoRet( &Jx_arr(lo.x+i_new-1+i, lo.y+k_new-1+k, 0, 0), sdxi);
#if defined(WARPX_DIM_RZ)
                    Complex xy_mid = xy_mid0; // Throughout the following loop, xy_mid takes the value e^{i m theta}
                    for (int imode=1 ; imode < n_rz_azimuthal_modes ; imode++) {
                        // The factor 2 comes from the normalization of the modes
                        const Complex djr_cmplx = 2._rt *sdxi*xy_mid;
                        amrex::Gpu::Atomic::AddNoRet( &Jx_arr(lo.x+i_new-1+i, lo.y+k_new-1+k, 0, 2*imode-1), djr_cmplx.real());
                        amrex::Gpu::Atomic::AddNoRet( &Jx_arr(lo.x+i_new-1+i, lo.y+k_new-1+k, 0, 2*imode), djr_cmplx.imag());
                        xy_mid = xy_mid*xy_mid0;
                    }
#endif
                }
            }
            for (int k=dkl; k<=depos_order+2-dku; k++) {
                for (int i=dil; i<=depos_order+2-diu; i++) {
                    Real const sdyj = wq*vy*invvol*(
                        one_third*(sx_new[i]*sz_new[k] + sx_old[i]*sz_old[k])
                       +one_sixth*(sx_new[i]*sz_old[k] + sx_old[i]*sz_new[k]));
                    amrex::Gpu::Atomic::AddNoRet( &Jy_arr(lo.x+i_new-1+i, lo.y+k_new-1+k, 0, 0), sdyj);
#if defined(WARPX_DIM_RZ)
                    Complex xy_new = xy_new0;
                    Complex xy_mid = xy_mid0;
                    Complex xy_old = xy_old0;
                    // Throughout the following loop, xy_ takes the value e^{i m theta_}
                    for (int imode=1 ; imode < n_rz_azimuthal_modes ; imode++) {
                        // The factor 2 comes from the normalization of the modes
                        // The minus sign comes from the different convention with respect to Davidson et al.
                        const Complex djt_cmplx = -2._rt * I*(i_new-1 + i + xmin*dxi)*wq*invdtdx/(amrex::Real)imode
                                                  *(Complex(sx_new[i]*sz_new[k], 0._rt)*(xy_new - xy_mid)
                                                  + Complex(sx_old[i]*sz_old[k], 0._rt)*(xy_mid - xy_old));
                        amrex::Gpu::Atomic::AddNoRet( &Jy_arr(lo.x+i_new-1+i, lo.y+k_new-1+k, 0, 2*imode-1), djt_cmplx.real());
                        amrex::Gpu::Atomic::AddNoRet( &Jy_arr(lo.x+i_new-1+i, lo.y+k_new-1+k, 0, 2*imode), djt_cmplx.imag());
                        xy_new = xy_new*xy_new0;
                        xy_mid = xy_mid*xy_mid0;
                        xy_old = xy_old*xy_old0;
                    }
#endif
                }
            }
            for (int i=dil; i<=depos_order+2-diu; i++) {
                Real sdzk = 0._rt;
                for (int k=dkl; k<=depos_order+1-dku; k++) {
                    sdzk += wqz*(sz_old[k] - sz_new[k])*0.5_rt*(sx_new[i] + sx_old[i]);
                    amrex::Gpu::Atomic::AddNoRet( &Jz_arr(lo.x+i_new-1+i, lo.y+k_new-1+k, 0, 0), sdzk);
#if defined(WARPX_DIM_RZ)
                    Complex xy_mid = xy_mid0; // Throughout the following loop, xy_mid takes the value e^{i m theta}
                    for (int imode=1 ; imode < n_rz_azimuthal_modes ; imode++) {
                        // The factor 2 comes from the normalization of the modes
                        const Complex djz_cmplx = 2._rt * sdzk * xy_mid;
                        amrex::Gpu::Atomic::AddNoRet( &Jz_arr(lo.x+i_new-1+i, lo.y+k_new-1+k, 0, 2*imode-1), djz_cmplx.real());
                        amrex::Gpu::Atomic::AddNoRet( &Jz_arr(lo.x+i_new-1+i, lo.y+k_new-1+k, 0, 2*imode), djz_cmplx.imag());
                        xy_mid = xy_mid*xy_mid0;
                    }
#endif
                }
            }
#elif defined(WARPX_DIM_1D_Z)

            for (int k=dkl; k<=depos_order+2-dku; k++) {
                amrex::Real const sdxi = wq*vx*invvol*0.5_rt*(sz_old[k] + sz_new[k]);
                amrex::Gpu::Atomic::AddNoRet( &Jx_arr(lo.x+k_new-1+k, 0, 0, 0), sdxi);
            }
            for (int k=dkl; k<=depos_order+2-dku; k++) {
                amrex::Real const sdyj = wq*vy*invvol*0.5_rt*(sz_old[k] + sz_new[k]);
                amrex::Gpu::Atomic::AddNoRet( &Jy_arr(lo.x+k_new-1+k, 0, 0, 0), sdyj);
            }
            amrex::Real sdzk = 0._rt;
            for (int k=dkl; k<=depos_order+1-dku; k++) {
                sdzk += wqz*(sz_old[k] - sz_new[k]);
                amrex::Gpu::Atomic::AddNoRet( &Jz_arr(lo.x+k_new-1+k, 0, 0, 0), sdzk);
            }
#endif
        }
    );
#if defined(WARPX_USE_GPUCLOCK)
    if(cost && load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock) {
        amrex::Gpu::streamSynchronize();
        *cost += *cost_real;
        amrex::The_Managed_Arena()->free(cost_real);
    }
#endif
}

/**
 * \brief Vay current deposition
 * (<a href="https://doi.org/10.1016/j.jcp.2013.03.010"> Vay et al, 2013</a>)
 * for thread \c thread_num: deposit \c D in real space and store the result in
 * \c Dx_fab, \c Dy_fab, \c Dz_fab
 *
 * \tparam depos_order  deposition order
 * \param[in] GetPosition  Functor that returns the particle position
 * \param[in] wp           Pointer to array of particle weights
 * \param[in] uxp,uyp,uzp  Pointer to arrays of particle momentum along \c x
 * \param[in] ion_lev      Pointer to array of particle ionization level. This is
                           required to have the charge of each macroparticle since \c q
                           is a scalar. For non-ionizable species, \c ion_lev is \c null
 * \param[in,out] Dx_fab,Dy_fab,Dz_fab FArrayBox of Vay current density, either full array or tile
 * \param[in] np_to_depose Number of particles for which current is deposited
 * \param[in] dt           Time step for particle level
 * \param[in] relative_time Time at which to deposit D, relative to the time of the
 *                          current positions of the particles. When different than 0,
 *                          the particle position will be temporarily modified to match
 *                          the time of the deposition.
 * \param[in] dx           3D cell size
 * \param[in] xyzmin       3D lower bounds of physical domain
 * \param[in] lo           Dimension-agnostic lower bounds of index domain
 * \param[in] q            Species charge
 * \param[in] n_rz_azimuthal_modes Number of azimuthal modes in RZ geometry
 * \param[in,out] cost     Pointer to (load balancing) cost corresponding to box where
                           present particles deposit current
 * \param[in] load_balance_costs_update_algo Selected method for updating load balance costs
 */
template <int depos_order>
void doVayDepositionShapeN (const GetParticlePosition& GetPosition,
                            const amrex::ParticleReal* const wp,
                            const amrex::ParticleReal* const uxp,
                            const amrex::ParticleReal* const uyp,
                            const amrex::ParticleReal* const uzp,
                            const int* const ion_lev,
                            amrex::FArrayBox& Dx_fab,
                            amrex::FArrayBox& Dy_fab,
                            amrex::FArrayBox& Dz_fab,
                            long np_to_depose,
                            amrex::Real dt,
                            amrex::Real relative_time,
                            const std::array<amrex::Real,3>& dx,
                            const std::array<amrex::Real,3>& xyzmin,
                            amrex::Dim3 lo,
                            amrex::Real q,
                            int n_rz_azimuthal_modes,
                            amrex::Real* cost,
                            long load_balance_costs_update_algo)
{
    using namespace amrex::literals;

#if defined(WARPX_DIM_RZ)
    amrex::ignore_unused(GetPosition,
        wp, uxp, uyp, uzp, ion_lev, Dx_fab, Dy_fab, Dz_fab,
        np_to_depose, dt, relative_time, dx, xyzmin, lo, q, n_rz_azimuthal_modes);
    WARPX_ABORT_WITH_MESSAGE("Vay deposition not implemented in RZ geometry");
#endif

#if defined(WARPX_DIM_1D_Z)
    amrex::ignore_unused(GetPosition,
        wp, uxp, uyp, uzp, ion_lev, Dx_fab, Dy_fab, Dz_fab,
        np_to_depose, dt, relative_time, dx, xyzmin, lo, q, n_rz_azimuthal_modes);
    WARPX_ABORT_WITH_MESSAGE("Vay deposition not implemented in cartesian 1D geometry");
#endif

#if !defined(WARPX_USE_GPUCLOCK)
    amrex::ignore_unused(cost, load_balance_costs_update_algo);
#endif

#if !(defined WARPX_DIM_RZ || defined WARPX_DIM_1D_Z)
    amrex::ignore_unused(n_rz_azimuthal_modes);

    // If ion_lev is a null pointer, then do_ionization=0, else do_ionization=1
    const bool do_ionization = ion_lev;

    // Inverse cell volume in each direction
    const amrex::Real dxi = 1._rt / dx[0];
    const amrex::Real dzi = 1._rt / dx[2];
#if defined(WARPX_DIM_3D)
    const amrex::Real dyi = 1._rt / dx[1];
#endif

    // Inverse of time step
    const amrex::Real invdt = 1._rt / dt;

    // Total inverse cell volume
#if   defined(WARPX_DIM_XZ)
    const amrex::Real invvol = dxi * dzi;
#elif defined(WARPX_DIM_3D)
    const amrex::Real invvol = dxi * dyi * dzi;
#endif

    // Lower bound of physical domain in each direction
    const amrex::Real xmin = xyzmin[0];
    const amrex::Real zmin = xyzmin[2];
#if defined(WARPX_DIM_3D)
    const amrex::Real ymin = xyzmin[1];
#endif

    // Allocate temporary arrays
#if defined(WARPX_DIM_3D)
    AMREX_ALWAYS_ASSERT(Dx_fab.box() == Dy_fab.box() && Dx_fab.box() == Dz_fab.box());
    amrex::FArrayBox temp_fab{Dx_fab.box(), 4};
#elif defined(WARPX_DIM_XZ)
    AMREX_ALWAYS_ASSERT(Dx_fab.box() == Dz_fab.box());
    amrex::FArrayBox temp_fab{Dx_fab.box(), 2};
#endif
    temp_fab.setVal<amrex::RunOn::Device>(0._rt);
    amrex::Array4<amrex::Real> const& temp_arr = temp_fab.array();

    // Inverse of light speed squared
    const amrex::Real invcsq = 1._rt / (PhysConst::c * PhysConst::c);

    // Arrays where D will be stored
    amrex::Array4<amrex::Real> const& Dx_arr = Dx_fab.array();
    amrex::Array4<amrex::Real> const& Dy_arr = Dy_fab.array();
    amrex::Array4<amrex::Real> const& Dz_arr = Dz_fab.array();

    // Loop over particles and deposit (Dx,Dy,Dz) into Dx_fab, Dy_fab and Dz_fab
#if defined(WARPX_USE_GPUCLOCK)
    amrex::Real* cost_real = nullptr;
    if( load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock) {
        cost_real = (amrex::Real *) amrex::The_Managed_Arena()->alloc(sizeof(amrex::Real));
        *cost_real = 0._rt;
    }
#endif
    amrex::ParallelFor(np_to_depose, [=] AMREX_GPU_DEVICE (long ip)
    {
#if defined(WARPX_USE_GPUCLOCK)
        const auto KernelTimer = ablastr::parallelization::KernelTimer(
            cost && (load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock),
            cost_real);
#endif

        // Inverse of Lorentz factor gamma
        const amrex::Real invgam = 1._rt / std::sqrt(1._rt + uxp[ip] * uxp[ip] * invcsq
                                                           + uyp[ip] * uyp[ip] * invcsq
                                                           + uzp[ip] * uzp[ip] * invcsq);
        // Product of particle charges and weights
        amrex::Real wq = q * wp[ip];
        if (do_ionization) wq *= ion_lev[ip];

        // Current particle positions (in physical units)
        amrex::ParticleReal xp, yp, zp;
        GetPosition(ip, xp, yp, zp);

        // Particle velocities
        const amrex::Real vx = uxp[ip] * invgam;
        const amrex::Real vy = uyp[ip] * invgam;
        const amrex::Real vz = uzp[ip] * invgam;

        // Modify the particle position to match the time of the deposition
        xp += relative_time * vx;
        yp += relative_time * vy;
        zp += relative_time * vz;

        // Particle current densities
#if defined(WARPX_DIM_XZ)
        const amrex::Real wqy = wq * vy * invvol;
#endif

        // Current and old particle positions in grid units
        // Keep these double to avoid bug in single precision.
        double const x_new = (xp - xmin + 0.5_rt*dt*vx) * dxi;
        double const x_old = (xp - xmin - 0.5_rt*dt*vx) * dxi;
#if defined(WARPX_DIM_3D)
        // Keep these double to avoid bug in single precision.
        double const y_new = (yp - ymin + 0.5_rt*dt*vy) * dyi;
        double const y_old = (yp - ymin - 0.5_rt*dt*vy) * dyi;
#endif
        // Keep these double to avoid bug in single precision.
        double const z_new = (zp - zmin + 0.5_rt*dt*vz) * dzi;
        double const z_old = (zp - zmin - 0.5_rt*dt*vz) * dzi;

        // Shape factor arrays for current and old positions (nodal)
        // Keep these double to avoid bug in single precision.
        double sx_new[depos_order+1] = {0.};
        double sx_old[depos_order+1] = {0.};
#if defined(WARPX_DIM_3D)
        // Keep these double to avoid bug in single precision.
        double sy_new[depos_order+1] = {0.};
        double sy_old[depos_order+1] = {0.};
#endif
        // Keep these double to avoid bug in single precision.
        double sz_new[depos_order+1] = {0.};
        double sz_old[depos_order+1] = {0.};

        // Compute shape factors for current positions

        // i_new leftmost grid point in x that the particle touches
        // sx_new shape factor along x for the centering of each current
        Compute_shape_factor< depos_order > const compute_shape_factor;
        const int i_new = compute_shape_factor(sx_new, x_new);
#if defined(WARPX_DIM_3D)
        // j_new leftmost grid point in y that the particle touches
        // sy_new shape factor along y for the centering of each current
        const int j_new = compute_shape_factor(sy_new, y_new);
#endif
        // k_new leftmost grid point in z that the particle touches
        // sz_new shape factor along z for the centering of each current
        const int k_new = compute_shape_factor(sz_new, z_new);

        // Compute shape factors for old positions

        // i_old leftmost grid point in x that the particle touches
        // sx_old shape factor along x for the centering of each current
        const int i_old = compute_shape_factor(sx_old, x_old);
#if defined(WARPX_DIM_3D)
        // j_old leftmost grid point in y that the particle touches
        // sy_old shape factor along y for the centering of each current
        const int j_old = compute_shape_factor(sy_old, y_old);
#endif
        // k_old leftmost grid point in z that the particle touches
        // sz_old shape factor along z for the centering of each current
        const int k_old = compute_shape_factor(sz_old, z_old);

        // Deposit current into Dx_arr, Dy_arr and Dz_arr
#if defined(WARPX_DIM_XZ)

        for (int k=0; k<=depos_order; k++) {
            for (int i=0; i<=depos_order; i++) {

                // Re-casting sx_new and sz_new from double to amrex::Real so that
                // Atomic::Add has consistent types in its argument
                auto const sxn_szn = static_cast<amrex::Real>(sx_new[i] * sz_new[k]);
                auto const sxo_szn = static_cast<amrex::Real>(sx_old[i] * sz_new[k]);
                auto const sxn_szo = static_cast<amrex::Real>(sx_new[i] * sz_old[k]);
                auto const sxo_szo = static_cast<amrex::Real>(sx_old[i] * sz_old[k]);

                if (i_new == i_old && k_new == k_old) {
                    // temp arrays for Dx and Dz
                    amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + k_new + k, 0, 0),
                        wq * invvol * invdt * (sxn_szn - sxo_szo));

                    amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + k_new + k, 0, 1),
                        wq * invvol * invdt * (sxn_szo - sxo_szn));

                    // Dy
                    amrex::Gpu::Atomic::AddNoRet(&Dy_arr(lo.x + i_new + i, lo.y + k_new + k, 0, 0),
                        wqy * 0.25_rt * (sxn_szn + sxn_szo + sxo_szn + sxo_szo));
                } else {
                    // temp arrays for Dx and Dz
                    amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + k_new + k, 0, 0),
                        wq * invvol * invdt * sxn_szn);

                    amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_old + i, lo.y + k_old + k, 0, 0),
                        - wq * invvol * invdt * sxo_szo);

                    amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + k_old + k, 0, 1),
                        wq * invvol * invdt * sxn_szo);

                    amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_old + i, lo.y + k_new + k, 0, 1),
                        - wq * invvol * invdt * sxo_szn);

                    // Dy
                    amrex::Gpu::Atomic::AddNoRet(&Dy_arr(lo.x + i_new + i, lo.y + k_new + k, 0, 0),
                        wqy * 0.25_rt * sxn_szn);

                    amrex::Gpu::Atomic::AddNoRet(&Dy_arr(lo.x + i_new + i, lo.y + k_old + k, 0, 0),
                        wqy * 0.25_rt * sxn_szo);

                    amrex::Gpu::Atomic::AddNoRet(&Dy_arr(lo.x + i_old + i, lo.y + k_new + k, 0, 0),
                        wqy * 0.25_rt * sxo_szn);

                    amrex::Gpu::Atomic::AddNoRet(&Dy_arr(lo.x + i_old + i, lo.y + k_old + k, 0, 0),
                        wqy * 0.25_rt * sxo_szo);
                }

            }
        }

#elif defined(WARPX_DIM_3D)

        for (int k=0; k<=depos_order; k++) {
            for (int j=0; j<=depos_order; j++) {

                auto const syn_szn = static_cast<amrex::Real>(sy_new[j] * sz_new[k]);
                auto const syo_szn = static_cast<amrex::Real>(sy_old[j] * sz_new[k]);
                auto const syn_szo = static_cast<amrex::Real>(sy_new[j] * sz_old[k]);
                auto const syo_szo = static_cast<amrex::Real>(sy_old[j] * sz_old[k]);

                for (int i=0; i<=depos_order; i++) {

                    auto const sxn_syn_szn = static_cast<amrex::Real>(sx_new[i]) * syn_szn;
                    auto const sxo_syn_szn = static_cast<amrex::Real>(sx_old[i]) * syn_szn;
                    auto const sxn_syo_szn = static_cast<amrex::Real>(sx_new[i]) * syo_szn;
                    auto const sxo_syo_szn = static_cast<amrex::Real>(sx_old[i]) * syo_szn;
                    auto const sxn_syn_szo = static_cast<amrex::Real>(sx_new[i]) * syn_szo;
                    auto const sxo_syn_szo = static_cast<amrex::Real>(sx_old[i]) * syn_szo;
                    auto const sxn_syo_szo = static_cast<amrex::Real>(sx_new[i]) * syo_szo;
                    auto const sxo_syo_szo = static_cast<amrex::Real>(sx_old[i]) * syo_szo;

                    if (i_new == i_old && j_new == j_old && k_new == k_old) {
                        // temp arrays for Dx, Dy and Dz
                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + j_new + j, lo.z + k_new + k, 0),
                            wq * invvol * invdt * (sxn_syn_szn - sxo_syo_szo));

                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + j_new + j, lo.z + k_new + k, 1),
                            wq * invvol * invdt * (sxn_syn_szo - sxo_syo_szn));

                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + j_new + j, lo.z + k_new + k, 2),
                            wq * invvol * invdt * (sxn_syo_szn - sxo_syn_szo));

                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + j_new + j, lo.z + k_new + k, 3),
                            wq * invvol * invdt * (sxo_syn_szn - sxn_syo_szo));
                    } else {
                        // temp arrays for Dx, Dy and Dz
                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + j_new + j, lo.z + k_new + k, 0),
                            wq * invvol * invdt * sxn_syn_szn);

                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_old + i, lo.y + j_old + j, lo.z + k_old + k, 0),
                            - wq * invvol * invdt * sxo_syo_szo);

                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + j_new + j, lo.z + k_old + k, 1),
                            wq * invvol * invdt * sxn_syn_szo);

                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_old + i, lo.y + j_old + j, lo.z + k_new + k, 1),
                            - wq * invvol * invdt * sxo_syo_szn);

                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + j_old + j, lo.z + k_new + k, 2),
                            wq * invvol * invdt * sxn_syo_szn);

                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_old + i, lo.y + j_new + j, lo.z + k_old + k, 2),
                            - wq * invvol * invdt * sxo_syn_szo);

                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_old + i, lo.y + j_new + j, lo.z + k_new + k, 3),
                            wq * invvol * invdt * sxo_syn_szn);

                        amrex::Gpu::Atomic::AddNoRet(&temp_arr(lo.x + i_new + i, lo.y + j_old + j, lo.z + k_old + k, 3),
                            - wq * invvol * invdt * sxn_syo_szo);
                    }
                }
            }
        }
#endif
    } );

#if defined(WARPX_DIM_3D)
    amrex::ParallelFor(Dx_fab.box(), [=] AMREX_GPU_DEVICE (int i, int j, int k) noexcept
    {
        const amrex::Real t_a = temp_arr(i,j,k,0);
        const amrex::Real t_b = temp_arr(i,j,k,1);
        const amrex::Real t_c = temp_arr(i,j,k,2);
        const amrex::Real t_d = temp_arr(i,j,k,3);
        Dx_arr(i,j,k) += (1._rt/6._rt)*(2_rt*t_a       + t_b       + t_c - 2._rt*t_d);
        Dy_arr(i,j,k) += (1._rt/6._rt)*(2_rt*t_a       + t_b - 2._rt*t_c       + t_d);
        Dz_arr(i,j,k) += (1._rt/6._rt)*(2_rt*t_a - 2._rt*t_b       + t_c       + t_d);
    });
#elif defined(WARPX_DIM_XZ)
    amrex::ParallelFor(Dx_fab.box(), [=] AMREX_GPU_DEVICE (int i, int j, int) noexcept
    {
        const amrex::Real t_a = temp_arr(i,j,0,0);
        const amrex::Real t_b = temp_arr(i,j,0,1);
        Dx_arr(i,j,0) += (0.5_rt)*(t_a + t_b);
        Dz_arr(i,j,0) += (0.5_rt)*(t_a - t_b);
    });
#endif
    // Synchronize so that temp_fab can be safely deallocated in its destructor
    amrex::Gpu::streamSynchronize();


#   if defined(WARPX_USE_GPUCLOCK)
    if( load_balance_costs_update_algo == LoadBalanceCostsUpdateAlgo::GpuClock) {
        amrex::Gpu::streamSynchronize();
        *cost += *cost_real;
        amrex::The_Managed_Arena()->free(cost_real);
    }
#   endif
#endif // #if !(defined WARPX_DIM_RZ || defined WARPX_DIM_1D_Z)
}
#endif // CURRENTDEPOSITION_H_