2 * This file is part of the GROMACS molecular simulation package.
4 * Copyright (c) 1991-2000, University of Groningen, The Netherlands.
5 * Copyright (c) 2001-2012, The GROMACS development team.
6 * Copyright (c) 2013-2019,2020, by the GROMACS development team, led by
7 * Mark Abraham, David van der Spoel, Berk Hess, and Erik Lindahl,
8 * and including many others, as listed in the AUTHORS file in the
9 * top-level source directory and at http://www.gromacs.org.
11 * GROMACS is free software; you can redistribute it and/or
12 * modify it under the terms of the GNU Lesser General Public License
13 * as published by the Free Software Foundation; either version 2.1
14 * of the License, or (at your option) any later version.
16 * GROMACS is distributed in the hope that it will be useful,
17 * but WITHOUT ANY WARRANTY; without even the implied warranty of
18 * MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the GNU
19 * Lesser General Public License for more details.
21 * You should have received a copy of the GNU Lesser General Public
22 * License along with GROMACS; if not, see
23 * http://www.gnu.org/licenses, or write to the Free Software Foundation,
24 * Inc., 51 Franklin Street, Fifth Floor, Boston, MA 02110-1301 USA.
26 * If you want to redistribute modifications to GROMACS, please
27 * consider that scientific software is very special. Version
28 * control is crucial - bugs must be traceable. We will be happy to
29 * consider code for inclusion in the official distribution, but
30 * derived work must not be called official GROMACS. Details are found
31 * in the README & COPYING files - if they are missing, get the
32 * official version at http://www.gromacs.org.
34 * To help us fund GROMACS development, we humbly ask that you cite
35 * the research papers on the package. Check out http://www.gromacs.org.
40 * Data types used internally in the nbnxn_cuda module.
42 * \author Szilárd Páll <pall.szilard@gmail.com>
43 * \ingroup module_nbnxm
46 #ifndef NBNXM_CUDA_TYPES_H
47 #define NBNXM_CUDA_TYPES_H
49 #include "gromacs/gpu_utils/cuda_arch_utils.cuh"
50 #include "gromacs/gpu_utils/cudautils.cuh"
51 #include "gromacs/gpu_utils/devicebuffer.h"
52 #include "gromacs/gpu_utils/devicebuffer_datatype.h"
53 #include "gromacs/gpu_utils/gputraits.cuh"
54 #include "gromacs/mdtypes/interaction_const.h"
55 #include "gromacs/nbnxm/gpu_types_common.h"
56 #include "gromacs/nbnxm/nbnxm.h"
57 #include "gromacs/nbnxm/pairlist.h"
58 #include "gromacs/timing/gpu_timing.h"
59 #include "gromacs/utility/enumerationhelpers.h"
61 /*! \brief Macro definining default for the prune kernel's j4 processing concurrency.
63 * The GMX_NBNXN_PRUNE_KERNEL_J4_CONCURRENCY macro allows compile-time override.
65 #ifndef GMX_NBNXN_PRUNE_KERNEL_J4_CONCURRENCY
66 # define GMX_NBNXN_PRUNE_KERNEL_J4_CONCURRENCY 4
68 /*! \brief Default for the prune kernel's j4 processing concurrency.
70 * Initialized using the #GMX_NBNXN_PRUNE_KERNEL_J4_CONCURRENCY macro which allows compile-time override.
72 const int c_cudaPruneKernelJ4Concurrency = GMX_NBNXN_PRUNE_KERNEL_J4_CONCURRENCY;
74 /* TODO: consider moving this to kernel_utils */
75 /* Convenience defines */
76 /*! \brief cluster size = number of atoms per cluster. */
77 static constexpr int c_clSize = c_nbnxnGpuClusterSize;
79 /* All structs prefixed with "cu_" hold data used in GPU calculations and
80 * are passed to the kernels, except cu_timers_t. */
82 typedef struct cu_atomdata cu_atomdata_t;
83 typedef struct cu_nbparam cu_nbparam_t;
88 * \brief Staging area for temporary data downloaded from the GPU.
90 * The energies/shift forces get downloaded here first, before getting added
91 * to the CPU-side aggregate values.
96 float* e_lj = nullptr;
97 //! electrostatic energy
98 float* e_el = nullptr;
100 float3* fshift = nullptr;
104 * \brief Nonbonded atom data - both inputs and outputs.
110 //! number of local atoms
112 //! allocation size for the atom data (xq, f)
115 //! atom coordinates + charges, size natoms
116 DeviceBuffer<float4> xq;
117 //! force output array, size natoms
118 DeviceBuffer<float3> f;
120 //! LJ energy output, size 1
121 DeviceBuffer<float> e_lj;
122 //! Electrostatics energy input, size 1
123 DeviceBuffer<float> e_el;
126 DeviceBuffer<float3> fshift;
128 //! number of atom types
130 //! atom type indices, size natoms
131 DeviceBuffer<int> atom_types;
132 //! sqrt(c6),sqrt(c12) size natoms
133 DeviceBuffer<float2> lj_comb;
136 DeviceBuffer<float3> shift_vec;
137 //! true if the shift vector has been uploaded
138 bool bShiftVecUploaded;
142 * \brief Parameters required for the CUDA nonbonded calculations.
147 //! type of electrostatics, takes values from #eelType
149 //! type of VdW impl., takes values from #evdwType
152 //! charge multiplication factor
154 //! Reaction-field/plain cutoff electrostatics const.
156 //! Reaction-field electrostatics constant
158 //! Ewald/PME parameter
160 //! Ewald/PME correction term substracted from the direct-space potential
162 //! LJ-Ewald/PME correction term added to the correction potential
164 //! LJ-Ewald/PME coefficient
167 //! Coulomb cut-off squared
170 //! VdW cut-off squared
172 //! VdW switched cut-off
174 //! Full, outer pair-list cut-off squared
176 //! Inner, dynamic pruned pair-list cut-off squared
178 //! True if we use dynamic pair-list pruning
179 bool useDynamicPruning;
181 //! VdW shift dispersion constants
182 shift_consts_t dispersion_shift;
183 //! VdW shift repulsion constants
184 shift_consts_t repulsion_shift;
185 //! VdW switch constants
186 switch_consts_t vdw_switch;
188 /* LJ non-bonded parameters - accessed through texture memory */
189 //! nonbonded parameter table with C6/C12 pairs per atom type-pair, 2*ntype^2 elements
191 //! texture object bound to nbfp
192 cudaTextureObject_t nbfp_texobj;
193 //! nonbonded parameter table per atom type, 2*ntype elements
195 //! texture object bound to nbfp_texobj
196 cudaTextureObject_t nbfp_comb_texobj;
198 /* Ewald Coulomb force table data - accessed through texture memory */
199 //! table scale/spacing
200 float coulomb_tab_scale;
201 //! pointer to the table in the device memory
203 //! texture object bound to coulomb_tab
204 cudaTextureObject_t coulomb_tab_texobj;
208 * \brief Pair list data.
210 using cu_plist_t = Nbnxm::gpu_plist;
213 * \brief Typedef of actual timer type.
215 typedef struct Nbnxm::gpu_timers_t cu_timers_t;
217 class GpuEventSynchronizer;
220 * \brief Main data structure for CUDA nonbonded force calculations.
224 /*! \brief GPU device context.
226 * \todo Make it constant reference, once NbnxmGpu is a proper class.
228 const DeviceContext* deviceContext_;
229 /*! \brief true if doing both local/non-local NB work on GPU */
230 bool bUseTwoStreams = false;
231 /*! \brief atom data */
232 cu_atomdata_t* atdat = nullptr;
233 /*! \brief f buf ops cell index mapping */
235 /*! \brief number of indices in cell buffer */
237 /*! \brief number of indices allocated in cell buffer */
239 /*! \brief array of atom indices */
240 int* atomIndices = nullptr;
241 /*! \brief size of atom indices */
242 int atomIndicesSize = 0;
243 /*! \brief size of atom indices allocated in device buffer */
244 int atomIndicesSize_alloc = 0;
245 /*! \brief x buf ops num of atoms */
246 int* cxy_na = nullptr;
247 /*! \brief number of elements in cxy_na */
249 /*! \brief number of elements allocated allocated in device buffer */
250 int ncxy_na_alloc = 0;
251 /*! \brief x buf ops cell index mapping */
252 int* cxy_ind = nullptr;
253 /*! \brief number of elements in cxy_ind */
255 /*! \brief number of elements allocated allocated in device buffer */
256 int ncxy_ind_alloc = 0;
257 /*! \brief parameters required for the non-bonded calc. */
258 cu_nbparam_t* nbparam = nullptr;
259 /*! \brief pair-list data structures (local and non-local) */
260 gmx::EnumerationArray<Nbnxm::InteractionLocality, cu_plist_t*> plist = { { nullptr } };
261 /*! \brief staging area where fshift/energies get downloaded */
263 /*! \brief local and non-local GPU streams */
264 gmx::EnumerationArray<Nbnxm::InteractionLocality, const DeviceStream*> deviceStreams;
266 /*! \brief Events used for synchronization */
268 /*! \brief Event triggered when the non-local non-bonded
269 * kernel is done (and the local transfer can proceed) */
270 cudaEvent_t nonlocal_done = nullptr;
271 /*! \brief Event triggered when the tasks issued in the local
272 * stream that need to precede the non-local force or buffer
273 * operation calculations are done (e.g. f buffer 0-ing, local
274 * x/q H2D, buffer op initialization in local stream that is
275 * required also by nonlocal stream ) */
276 cudaEvent_t misc_ops_and_local_H2D_done = nullptr;
279 /*! \brief True if there is work for the current domain in the
280 * respective locality.
282 * This includes local/nonlocal GPU work, either bonded or
283 * nonbonded, scheduled to be executed in the current
284 * domain. As long as bonded work is not split up into
285 * local/nonlocal, if there is bonded GPU work, both flags
287 gmx::EnumerationArray<Nbnxm::InteractionLocality, bool> haveWork = { { false } };
289 /*! \brief Pointer to event synchronizer triggered when the local
290 * GPU buffer ops / reduction is complete
292 * \note That the synchronizer is managed outside of this module
293 * in StatePropagatorDataGpu.
295 GpuEventSynchronizer* localFReductionDone = nullptr;
297 /*! \brief Event triggered when non-local coordinate buffer
298 * has been copied from device to host. */
299 GpuEventSynchronizer* xNonLocalCopyD2HDone = nullptr;
301 /* NOTE: With current CUDA versions (<=5.0) timing doesn't work with multiple
302 * concurrent streams, so we won't time if both l/nl work is done on GPUs.
303 * Timer init/uninit is still done even with timing off so only the condition
304 * setting bDoTime needs to be change if this CUDA "feature" gets fixed. */
305 /*! \brief True if event-based timing is enabled. */
306 bool bDoTime = false;
307 /*! \brief CUDA event-based timers. */
308 cu_timers_t* timers = nullptr;
309 /*! \brief Timing data. TODO: deprecate this and query timers for accumulated data instead */
310 gmx_wallclock_gpu_nbnxn_t* timings = nullptr;
313 #endif /* NBNXN_CUDA_TYPES_H */