Molecular modeling on GPUs

{{short description|Using graphics processing units for molecular simulations}}

Image:Hardware-accelerated-molecular-modeling.png)]]

Molecular modeling on GPU is the technique of using a graphics processing unit (GPU) for molecular simulations.{{cite journal | vauthors = Stone JE, Phillips JC, Freddolino PL, Hardy DJ, Trabuco LG, Schulten K | title = Accelerating molecular modeling applications with graphics processors | journal = Journal of Computational Chemistry | volume = 28 | issue = 16 | pages = 2618–2640 | date = December 2007 | pmid = 17894371 | doi = 10.1002/jcc.20829 | citeseerx = 10.1.1.466.3823 | s2cid = 15313533 }}

In 2007, Nvidia introduced video cards that could be used not only to show graphics but also for scientific calculations. These cards include many arithmetic units ({{as of|2016|lc=y}}, up to 3,584 in Tesla P100) working in parallel. Long before this event, the computational power of video cards was purely used to accelerate graphics calculations. The new features of these cards made it possible to develop parallel programs in a high-level application programming interface (API) named CUDA. This technology substantially simplified programming by enabling programs to be written in C/C++. More recently, OpenCL allows cross-platform GPU acceleration.

Quantum chemistry calculations{{cite journal | vauthors = Yasuda K | title = Accelerating Density Functional Calculations with Graphics Processing Unit | journal = Journal of Chemical Theory and Computation | volume = 4 | issue = 8 | pages = 1230–1236 | date = August 2008 | pmid = 26631699 | doi = 10.1021/ct8001046 }}{{cite journal | vauthors = Yasuda K | title = Two-electron integral evaluation on the graphics processor unit | journal = Journal of Computational Chemistry | volume = 29 | issue = 3 | pages = 334–342 | date = February 2008 | pmid = 17614340 | doi = 10.1002/jcc.20779 | citeseerx = 10.1.1.498.364 | s2cid = 8078401 }}{{cite journal | vauthors = Vogt L, Olivares-Amaya R, Kermes S, Shao Y, Amador-Bedolla C, Aspuru-Guzik A | title = Accelerating resolution-of-the-identity second-order Møller-Plesset quantum chemistry calculations with graphical processing units | journal = The Journal of Physical Chemistry A | volume = 112 | issue = 10 | pages = 2049–2057 | date = March 2008 | pmid = 18229900 | doi = 10.1021/jp0776762 | bibcode = 2008JPCA..112.2049V | s2cid = 4566211 | url = http://nrs.harvard.edu/urn-3:HUL.InstRepos:5344183 | url-access = subscription }}{{cite journal | vauthors = Ufimtsev IS, Martínez TJ | title = Quantum Chemistry on Graphical Processing Units. 1. Strategies for Two-Electron Integral Evaluation | journal = Journal of Chemical Theory and Computation | volume = 4 | issue = 2 | pages = 222–231 | date = February 2008 | pmid = 26620654 | doi = 10.1021/ct700268q | name-list-style = amp }}{{cite journal |doi= 10.1109/MCSE.2008.148 |title= Graphical Processing Units for Quantum Chemistry |author1=Ivan S. Ufimtsev |author2=Todd J. Martinez |name-list-style=amp |journal= Computing in Science & Engineering |volume= 10 |issue= 6 |year= 2008 |pages= 26–34|bibcode= 2008CSE....10f..26U |s2cid= 10225262 }}{{cite journal | vauthors = Tornai GJ, Ladjánszki I, Rák Á, Kis G, Cserey G | title = Calculation of Quantum Chemical Two-Electron Integrals by Applying Compiler Technology on GPU | journal = Journal of Chemical Theory and Computation | volume = 15 | issue = 10 | pages = 5319–5331 | date = October 2019 | pmid = 31503475 | doi = 10.1021/acs.jctc.9b00560 | s2cid = 202555796 | name-list-style = amp }} and molecular mechanics simulations{{cite journal |doi= 10.1016/j.jcp.2008.01.047 |title= General Purpose Molecular Dynamics Simulations Fully Implemented on Graphics Processing Units |author1=Joshua A. Anderson |author2=Chris D. Lorenz |author3=A. Travesset |journal= Journal of Computational Physics |volume= 227 |issue= 10 |year=2008 |pages= 5342–5359 |bibcode= 2008JCoPh.227.5342A|citeseerx= 10.1.1.552.2883 }}{{cite journal |title= GPU acceleration of cutoff pair potentials for molecular modeling applications. |author1=Christopher I. Rodrigues |author2=David J. Hardy |author3=John E. Stone |author4=Klaus Schulten |author5=Wen-Mei W. Hwu. |name-list-style=amp |journal= In CF'08: Proceedings of the 2008 Conference on Computing Frontiers, New York, NY, USA |year=2008 |pages= 273–282}}{{cite journal |doi= 10.1016/j.cpc.2011.01.009 |title= Highly accelerated simulations of glassy dynamics using GPUs: Caveats on limited floating-point precision |author1=Peter H. Colberg |author2=Felix Höfling |journal= Comput. Phys. Commun. |volume= 182 |issue= 5 |year= 2011 |pages= 1120–1129 |arxiv= 0912.3824 |bibcode= 2011CoPhC.182.1120C|s2cid= 7173093 }} (molecular modeling in terms of classical mechanics) are among beneficial applications of this technology. The video cards can accelerate the calculations tens of times, so a PC with such a card has the power similar to that of a cluster of workstations based on common processors.

GPU accelerated molecular modelling software

=Programs=

Abalone – Molecular Dynamics ([http://www.biomolecular-modeling.com/Abalone/Benchmarking.html Benchmark])
ACEMD on GPUs since [https://pubs.acs.org/doi/10.1021/ct9000685 2009] [https://www.acellera.com/products/molecular-dynamics-software-gpu-acemd/ Benchmark]
AMBER on GPUs [http://ambermd.org/gpus/ version]
[http://www.biomolecular-modeling.com/Products.html Ascalaph] on GPUs version – [http://www.biomolecular-modeling.com/Ascalaph/Ascalaph-Liquid.html Ascalaph Liquid GPU]
AutoDock – Molecular docking
BigDFT Ab initio program based on wavelet
[http://www.brianqc.com BrianQC] Quantum chemistry (HF and DFT) and molecular mechanics
[http://www.cresset-group.com/blaze/ Blaze] ligand-based virtual screening
CHARMM – Molecular dynamics [https://www.academiccharmm.org/]
CP2K Ab initio molecular dynamics
Desmond (software) on GPUs, workstations, and clusters
Firefly (formerly PC GAMESS)
[http://www.eyesopen.com/fastrocs FastROCS]
[http://gomc.eng.wayne.edu/ GOMC] – GPU Optimized Monte Carlo simulation engine
[http://gpiutmd.iut.ac.ir/ GPIUTMD] – Graphical processors for Many-Particle Dynamics
[https://github.com/pyscf/gpu4pyscf GPU4PySCF] – GPU accelerated plugin package for [https://github.com/pyscf/pyscf PySCF]
[https://gpumd.org/ GPUMD] - A light weight general-purpose molecular dynamics code
GROMACS on GPUs {{Cite journal| vauthors = Yousif RH |date=2020|title=Exploring the Molecular Interactions between Neoculin and the Human Sweet Taste Receptors through Computational Approaches|url=http://www.ukm.my/jsm/pdf_files/SM-PDF-49-3-2020/ARTIKEL%206.pdf|journal=Sains Malaysiana|volume=49|issue=3|pages=517–525|doi=10.17576/jsm-2020-4903-06|doi-access=free}}
[http://halmd.org/ HALMD] – Highly Accelerated Large-scale MD package
[https://glotzerlab.engin.umich.edu/hoomd-blue/ HOOMD-blue] {{Webarchive|url=https://web.archive.org/web/20111111003320/http://codeblue.umich.edu/hoomd-blue/index.html |date=2011-11-11 }} – Highly Optimized Object-oriented Many-particle Dynamics—Blue Edition
LAMMPS on GPUs version – [http://lammps.sandia.gov/doc/Section_accelerate.html lammps for accelerators]
[https://github.com/MALBECC/lio LIO] DFT-Based GPU optimized code - [https://github.com/MALBECC/lio]
Octopus has support for OpenCL.
[http://dna.physics.ox.ac.uk/ oxDNA] – DNA and RNA coarse-grained simulations on GPUs
[https://web.archive.org/web/20160121101500/http://www.pwmat.com/pwmat PWmat] – Plane-Wave Density Functional Theory simulations
[http://rumd.org RUMD] - Roskilde University Molecular Dynamics{{Cite journal| vauthors = Bailey N, Ingebrigtsen T, Hansen JS, Veldhorst A, Bøhling L, Lemarchand C, Olsen A, Bacher A, Costigliola L, Pedersen U, Larsen H | display-authors = 6 |date=2017-12-14|title=RUMD: A general purpose molecular dynamics package optimized to utilize GPU hardware down to a few thousand particles|journal=SciPost Physics|language=en|volume=3|issue=6|pages=038|doi=10.21468/SciPostPhys.3.6.038| bibcode = 2017ScPP....3...38B | s2cid = 43964588 |issn=2542-4653|doi-access=free|arxiv=1506.05094}}
TeraChem – Quantum chemistry and ab initio Molecular Dynamics
TINKER on GPUs.{{cite journal | vauthors = Harger M, Li D, Wang Z, Dalby K, Lagardère L, Piquemal JP, Ponder J, Ren P | display-authors = 6 | title = Tinker-OpenMM: Absolute and relative alchemical free energies using AMOEBA on GPUs | journal = Journal of Computational Chemistry | volume = 38 | issue = 23 | pages = 2047–2055 | date = September 2017 | pmid = 28600826 | pmc = 5539969 | doi = 10.1002/jcc.24853 }}
VMD & NAMD on GPUs [http://www.ks.uiuc.edu/Research/gpu/ versions]
YASARA runs MD simulations on all GPUs using OpenCL.

=API=

[http://brianqc.com BrianQC] – has an open C level API for quantum chemistry simulations on GPUs, provides GPU-accelerated version of Q-Chem and PSI
[http://simtk.org/home/openmm/ OpenMM] – an API for accelerating molecular dynamics on GPUs, v1.0 provides GPU-accelerated version of GROMACS
[http://mdcore.sourceforge.net mdcore] – an open-source platform-independent library for molecular dynamics simulations on modern shared-memory parallel architectures.