2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 2020,2021, by the GROMACS development team, led by
5 * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
6 * and including many others, as listed in the AUTHORS file in the
7 * top-level source directory and at http://www.gromacs.org.
9 * GROMACS is free software; you can redistribute it and/or
10 * modify it under the terms of the GNU Lesser General Public License
11 * as published by the Free Software Foundation; either version 2.1
12 * of the License, or (at your option) any later version.
14 * GROMACS is distributed in the hope that it will be useful,
15 * but WITHOUT ANY WARRANTY; without even the implied warranty of
16 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
17 * Lesser General Public License for more details.
19 * You should have received a copy of the GNU Lesser General Public
20 * License along with GROMACS; if not, see
21 * http://www.gnu.org/licenses, or write to the Free Software Foundation,
22 * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
24 * If you want to redistribute modifications to GROMACS, please
25 * consider that scientific software is very special. Version
26 * control is crucial - bugs must be traceable. We will be happy to
27 * consider code for inclusion in the official distribution, but
28 * derived work must not be called official GROMACS. Details are found
29 * in the README & COPYING files - if they are missing, get the
30 * official version at http://www.gromacs.org.
32 * To help us fund GROMACS development, we humbly ask that you cite
33 * the research papers on the package. Check out http://www.gromacs.org.
37 * \brief Declares the GPU Force Reduction
39 * \author Alan Gray <alang@nvidia.com>
41 * \ingroup module_mdlib
43 #ifndef GMX_MDLIB_GPUFORCEREDUCTION_IMPL_H
44 #define GMX_MDLIB_GPUFORCEREDUCTION_IMPL_H
46 #include "gromacs/gpu_utils/device_stream.h"
47 #include "gromacs/gpu_utils/devicebuffer_datatype.h"
48 #include "gromacs/math/vectypes.h"
50 #include "gpuforcereduction.h"
55 //! structure to hold cell information for any nbat-format forces
58 //! cell index mapping for any nbat-format forces
59 const int* cell = nullptr;
60 //! device copy of cell index mapping for any nbat-format forces
61 int* d_cell = nullptr;
62 //! number of atoms in cell array
64 //! number of atoms allocated in cell array
65 int cellSizeAlloc = -1;
68 class GpuForceReduction::Impl
72 /*! \brief Creates GPU force reduction object
74 * \param [in] deviceStream Stream to use for reduction
75 * \param [in] deviceContext GPU device context
76 * \param [in] wcycle The wallclock counter
78 Impl(const DeviceContext& deviceContext, const DeviceStream& deviceStreami, gmx_wallcycle* wcycle);
81 /*! \brief Register a nbnxm-format force to be reduced
83 * \param [in] forcePtr Pointer to force to be reduced
85 void registerNbnxmForce(DeviceBuffer<RVec> forcePtr);
87 /*! \brief Register a rvec-format force to be reduced
89 * \param [in] forcePtr Pointer to force to be reduced
91 void registerRvecForce(DeviceBuffer<RVec> forcePtr);
93 /*! \brief Add a dependency for this force reduction
95 * \param [in] dependency Dependency for this reduction
97 void addDependency(GpuEventSynchronizer* const dependency);
99 /*! \brief Reinitialize the GPU force reduction
101 * \param [in] baseForcePtr Pointer to force to be used as a base
102 * \param [in] numAtoms The number of atoms
103 * \param [in] cell Pointer to the cell array
104 * \param [in] atomStart The start atom for the reduction
105 * \param [in] accumulate Whether reduction should be accumulated
106 * \param [in] completionMarker Event to be marked when launch of reduction is complete
108 void reinit(float3* baseForcePtr,
110 ArrayRef<const int> cell,
112 const bool accumulate,
113 GpuEventSynchronizer* completionMarker = nullptr);
115 /*! \brief Execute the force reduction */
119 //! force to be used as a base for this reduction
120 float3* baseForce_ = nullptr;
125 //! whether reduction is accumulated into base force buffer
126 int accumulate_ = true;
127 //! cell information for any nbat-format forces
128 struct cellInfo cellInfo_;
129 //! GPU context object
130 const DeviceContext& deviceContext_;
131 //! list of dependencies
132 gmx::FixedCapacityVector<GpuEventSynchronizer*, 3> dependencyList_;
133 //! stream to be used for this reduction
134 const DeviceStream& deviceStream_;
135 //! Nbnxm force to be added in this reduction
136 DeviceBuffer<RVec> nbnxmForceToAdd_ = nullptr;
137 //! Rvec-format force to be added in this reduction
138 DeviceBuffer<RVec> rvecForceToAdd_ = nullptr;
139 //! event to be marked when redcution launch has been completed
140 GpuEventSynchronizer* completionMarker_ = nullptr;
141 //! The wallclock counter
142 gmx_wallcycle* wcycle_ = nullptr;