Anton 3: twenty microseconds of molecular dynamics simulation before lunch

Anton 3 is the newest member in a family of supercomputers specially designed for atomic-level simulation of molecules relevant to biology (e.g., DNA, proteins, and drug molecules). Anton 3 achieves order-of-magnitude improvements in time-to-solution over its predecessor, Anton 2 (the current state of the art), and is over 100-fold faster than any other currently available supercomputer, thereby enabling broad new avenues of research on critical questions in biology and drug discovery. This speedup means that a 512-node Anton 3 simulates a million atoms at over 100 microseconds per day. Furthermore, Anton 3 attains this performance while consuming an order of magnitude less energy per simulated microsecond than any other machine. Like its predecessors, Anton 3 was designed from the ground up around a new custom chip to best exploit the capabilities offered by new technologies. We present here the main architectural and algorithmic developments that were necessary to achieve such significant advances.

[1]  E. Lindahl,et al.  Heterogeneous Parallelization and Acceleration of Molecular Dynamics Simulations in GROMACS , 2020, The Journal of chemical physics.

[2]  ff19SB: Amino-acid specific protein backbone parameters trained against quantum mechanics energy surfaces in solution. , 2019, Journal of chemical theory and computation.

[3]  Luca Benini,et al.  On the Feasibility of FPGA Acceleration of Molecular Dynamics Simulations , 2018, ArXiv.

[4]  Stefano Piana,et al.  Demonstrating an Order-of-Magnitude Sampling Enhancement in Molecular Dynamics Simulations of Complex Protein Systems. , 2016, Journal of chemical theory and computation.

[5]  Ron O. Dror,et al.  Molecular Dynamics Simulation for All , 2018, Neuron.

[6]  Jianpeng Ma,et al.  CHARMM: The biomolecular simulation program , 2009, J. Comput. Chem..

[7]  Yi Wang,et al.  Scalable molecular dynamics on CPU and GPU architectures with NAMD. , 2020, The Journal of chemical physics.

[8]  M J Harvey,et al.  ACEMD: Accelerating Biomolecular Dynamics in the Microsecond Time Scale. , 2009, Journal of chemical theory and computation.

[9]  Mark E. Tuckerman,et al.  Reversible multiple time scale molecular dynamics , 1992 .

[10]  Makoto Taiji,et al.  MDGRAPE-4: a special-purpose computer system for molecular dynamics simulations , 2014, Philosophical Transactions of the Royal Society A: Mathematical, Physical and Engineering Sciences.

[11]  Yasuhiro Matsunaga,et al.  GENESIS 1.1: A hybrid‐parallel molecular dynamics simulator with enhanced sampling algorithms on multiple computational platforms , 2017, J. Comput. Chem..

[12]  John L. Klepeis,et al.  Millisecond-scale molecular dynamics simulations on Anton , 2009, Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis.

[13]  Jiayi Sheng,et al.  Fully Integrated On-FPGA Molecular Dynamics Simulations , 2019, ArXiv.

[14]  B. L. de Groot,et al.  CHARMM36m: an improved force field for folded and intrinsically disordered proteins , 2016, Nature Methods.

[15]  Robert D. Skeel,et al.  Corrected potential energy functions for constrained molecular dynamics , 2011 .

[16]  Shantenu Jha,et al.  AI-driven multiscale simulations illuminate mechanisms of SARS-CoV-2 spike dynamics , 2021, Int. J. High Perform. Comput. Appl..

[17]  K. Schulten,et al.  Physical properties of the HIV-1 capsid from all-atom molecular dynamics simulations , 2017, Nature Communications.

[18]  Pak Lui,et al.  Strong scaling of general-purpose molecular dynamics simulations on GPUs , 2014, Comput. Phys. Commun..

[19]  Martin C. Herbordt,et al.  Efficient particle-pair filtering for acceleration of molecular dynamics simulation , 2009, 2009 International Conference on Field Programmable Logic and Applications.

[20]  Sun UltraSPARC,et al.  A closer look at GPUs , 2008, Commun. ACM.

[21]  Albert C. Pan,et al.  Atomic-level characterization of protein–protein association , 2018, Proceedings of the National Academy of Sciences.

[22]  Alexander D. MacKerell,et al.  Further Optimization and Validation of the Classical Drude Polarizable Protein Force Field. , 2020, Journal of chemical theory and computation.

[23]  David E. Shaw,et al.  A fast, scalable method for the parallel evaluation of distance‐limited pairwise particle interactions , 2005, J. Comput. Chem..

[24]  John L. Klepeis,et al.  Anton, a special-purpose machine for molecular dynamics simulation , 2007, ISCA '07.

[25]  K. Lindorff-Larsen,et al.  Atomic-level description of ubiquitin folding , 2013, Proceedings of the National Academy of Sciences.

[26]  Paul Robustelli,et al.  Development of a force field for the simulation of single-chain proteins and protein-protein complexes. , 2020, Journal of chemical theory and computation.

[27]  Vijay S. Pande,et al.  OpenMM 7: Rapid development of high performance algorithms for molecular dynamics , 2016, bioRxiv.

[28]  J. P. Grossman,et al.  Millisecond-scale molecular dynamics simulations on Anton , 2009, Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis.

[29]  R. Dror,et al.  Systematic Validation of Protein Force Fields against Experimental Data , 2012, PloS one.

[30]  Berk Hess,et al.  Improving Efficiency of Large Time-Scale Molecular Dynamics Simulations of Hydrogen-Rich Systems , 1999 .

[31]  Martin Fechner,et al.  More bang for your buck: Improved use of GPU nodes for GROMACS 2018 , 2019, J. Comput. Chem..

[32]  G. Ciccotti,et al.  Numerical Integration of the Cartesian Equations of Motion of a System with Constraints: Molecular Dynamics of n-Alkanes , 1977 .

[33]  David E. Shaw,et al.  Overview of neutral territory methods for the parallel evaluation of pairwise particle interactions , 2005 .

[34]  J. P. Grossman,et al.  The u-series: A separable decomposition for electrostatics computation with improved accuracy. , 2020, The Journal of chemical physics.

[35]  Duncan Poole,et al.  Routine Microsecond Molecular Dynamics Simulations with AMBER on GPUs. 2. Explicit Solvent Particle Mesh Ewald. , 2013, Journal of chemical theory and computation.

[36]  Steve Plimpton,et al.  Fast parallel algorithms for short-range molecular dynamics , 1993 .

[37]  Sharon C. Glotzer,et al.  HOOMD-blue: A Python package for high-performance molecular dynamics and hard particle Monte Carlo simulations , 2013 .

[38]  T. Darden,et al.  Particle mesh Ewald: An N⋅log(N) method for Ewald sums in large systems , 1993 .

[39]  R. Dror,et al.  Gaussian split Ewald: A fast Ewald mesh method for molecular simulation. , 2005, The Journal of chemical physics.

[40]  Berk Hess,et al.  Improving efficiency of large time‐scale molecular dynamics simulations of hydrogen‐rich systems , 1999, Journal of computational chemistry.

[41]  J. P. Grossman,et al.  Anton 2: Raising the Bar for Performance and Programmability in a Special-Purpose Molecular Dynamics Supercomputer , 2014, SC14: International Conference for High Performance Computing, Networking, Storage and Analysis.

[42]  Robert M. Dirks,et al.  RNA force field with accuracy comparable to state-of-the-art protein force fields , 2018, Proceedings of the National Academy of Sciences.

[43]  Christopher G. Myers,et al.  Biomolecular Modeling and Simulation: A Prospering Multidisciplinary Field. , 2021, Annual review of biophysics.

[44]  Division on Earth.  COMMITTEE ON PROPOSAL EVALUATION FOR ALLOCATION OF SUPERCOMPUTING TIME FOR THE STUDY OF MOLECULAR DYNAMICS, THIRD ROUND , 2012 .

[45]  Rommie E. Amaro,et al.  A multiscale coarse-grained model of the SARS-CoV-2 virion , 2020, Biophysical Journal.

[46]  Carsten Kutzner,et al.  Tackling Exascale Software Challenges in Molecular Dynamics Simulations with GROMACS , 2015, EASC.