arXiv subjects

Danny Perez

Publications and source records attributed to Danny Perez.

49 records · Page 3Linked to original sources

arXiv2020

Rapid Exploration of Optimization Strategies on Advanced Architectures using TestSNAP and LAMMPS

The exascale race is at an end with the announcement of the Aurora and Frontier machines. This next generation of supercomputers utilize diverse hardware architectures to achieve their compute performance, providing an added onus on the performance portability of applications. An expanding fragmentation of programming models would provide a compounding optimization challenge were it not for the evolution of performance-portable frameworks, providing unified models for mapping abstract hierarchies of parallelism to diverse architectures. A solution to this challenge is the evolution of performance-portable frameworks, providing unified models for mapping abstract hierarchies of parallelism to diverse architectures. Kokkos is one such performance portable programming model for C++ applications, providing back-end implementations for each major HPC platform. Even with a performance portable framework, restructuring algorithms to expose higher degrees of parallelism is non-trivial. The Spectral Neighbor Analysis Potential (SNAP) is a machine-learned inter-atomic potential utilized in cutting-edge molecular dynamics simulations. Previous implementations of the SNAP calculation showed a downward trend in their performance relative to peak on newer-generation CPUs and low performance on GPUs. In this paper we describe the restructuring and optimization of SNAP as implemented in the Kokkos CUDA backend of the LAMMPS molecular dynamics package, benchmarked on NVIDIA GPUs. We identify novel patterns of hierarchical parallelism, facilitating a minimization of memory access overheads and pushing the implementation into a compute-saturated regime. Our implementation via Kokkos enables recompile-and-run efficiency on upcoming architectures. We find a $\sim$22x time-to-solution improvement relative to an existing implementation as measured on an NVIDIA Tesla V100-16GB for an important benchmark.

cs.DC↗

arXiv2020

Resource allocation for task-level speculative scientific applications: a proof of concept using Parallel Trajectory Splicing

The constant increase in parallelism available on large-scale distributed computers poses major scalability challenges to many scientific applications. A common strategy to improve scalability is to express the algorithm in terms of independent tasks that can be executed concurrently on a runtime system. In this manuscript, we consider a generalization of this approach where task-level speculation is allowed. In this context, a probability is attached to each task which corresponds to the likelihood that the product of the task will be consumed as part of the calculation. We consider the problem of optimal resource allocation to each of the possible tasks so as too maximize the expected overall computational throughput. The power of this approach is demonstrated by analyzing its application to Parallel Trajectory Splicing, a massively-parallel long-time-dynamics method for atomistic simulations.

cs.DC↗

arXiv2020

Automated calculation and convergence of defect transport tensors

Defect transport is a key process in materials science and catalysis, but as migration mechanisms are often too complex to enumerate a priori, calculation of transport tensors typically have no measure of convergence and require significant end user intervention. These two bottlenecks prevent high-throughput implementations essential to propagate model-form uncertainty from interatomic interactions to predictive simulations. In order to address these issues, we extend a massively parallel accelerated sampling scheme, autonomously controlled by Bayesian estimators of statewise sampling completeness, to build atomistic kinetic Monte Carlo models on a state space irreducible under exchange and space group symmetries. Focusing on isolated defects, we derive analytic expressions for defect transport tensors and provide a convergence metric by calculating the Kullback-Leiber divergence across the ensemble of diffusion processes consistent with the sampling uncertainty. The autonomy and efficacy of the method is demonstrated on surface trimers in tungsten and hexa-interstitials in magnesium oxide, both of which exhibit complex, correlated migration mechanisms.

physics.comp-ph↗

arXiv2020

Arbitrarily accurate representation of atomistic dynamics via Markov Renewal Processes

Atomistic simulations with methods such as molecular dynamics are extremely powerful tools to understand nanoscale dynamical behavior. The resulting trajectories, by the virtue of being embedded in a high-dimensional configuration space, can however be difficult to analyze and interpret. This makes low-dimensional representations, especially in terms of discrete jump processes, extremely valuable. This simplicity however usually comes at the cost of accuracy, as tractable representations often entail simplifying assumptions that are not guaranteed to be realized in practice. In this paper, we describe a discretization scheme for continuous trajectories that enables an arbitrarily accurate representation in terms of a Markov Renewal Process over a discrete state space. The accuracy of the model converges exponentially fast as a function of a continuous parameter that has the interpretation of a local correlation time of the dynamics.

cond-mat.stat-mech↗

arXiv2020

An Entropy-Maximization Approach to Automated Training Set Generation for Interatomic Potentials

Machine learning (ML)-based interatomic potentials are currently garnering a lot of attention as they strive to achieve the accuracy of electronic structure methods at the computational cost of empirical potentials. Given their generic functional forms, the transferability of these potentials is highly dependent on the quality of the training set, the generation of which is a highly labor-intensive activity. Good training sets should at once contain a very diverse set of configurations while avoiding redundancies that incur cost without providing benefits. We formalize these requirements in a local entropy maximization framework and propose an automated sampling scheme to sample from this objective function. We show that this approach generates much more diverse training sets than unbiased sampling and is competitive with hand-crafted training sets.

physics.comp-ph↗

arXiv2018

Self-optimized construction of transition rate matrices from accelerated atomistic simulations with Bayesian uncertainty quantification

A massively parallel method to build large transition rate matrices from temperature accelerated molecular dynamics trajectories is presented. Bayesian Markov model analysis is used to estimate the expected residence time in the known state space, providing crucial uncertainty quantification for higher scale simulation schemes such as kinetic Monte Carlo or cluster dynamics. The estimators are additionally used to optimize where exploration is performed and the degree of temperature ac- celeration on the fly, giving an autonomous, optimal procedure to explore the state space of complex systems. The method is tested against exactly solvable models and used to explore the dynamics of C15 interstitial defects in iron. Our uncertainty quantification scheme allows for accurate modeling of the evolution of these defects over timescales of several seconds.

physics.comp-ph↗

arXiv2016

Evidence for percolation diffusion of cations and material recovery in disordered pyrochlore from accelerated molecular dynamics simulations

We used classical and accelerated molecular dynamics simulations to characterize vacancy-mediated diffusion of cations in Gd$_2$Ti$_2$O$_7$ pyrochlore as a function of the disorder on the microsecond timescale. We find that cation vacancy diffusion is slow in materials with low levels of disorder. However, higher levels of disorder allow for fast cation diffusion, which is then also accompanied by fast antisite annihilation and ordering of the cations. The cation diffusivity is therefore not constant, but decreases as the material reorders. The results suggest that fast cation diffusion is triggered by the existence of a percolation network of antisites. This is in marked contrast with oxygen diffusion, which showed a smooth increase of the ionic diffusivity with increasing disorder in the same compound. The increase of the cation diffusivity with disorder is also contrary to observations from other complex oxides and disordered media models, suggesting a fundamentally different relation between disorder and mass transport. These results highlight the dynamic interplay between fast cation diffusion and the recovery of disorder and have important implications for understanding radiation damage evolution, sintering and aging, as well as diffusion in disorder oxides more generally.

cond-mat.mtrl-sci↗

arXiv2015

Accelerating the convergence of replica exchange simulations using Gibbs sampling and adaptive temperature sets

We recently introduced a novel replica-exchange scheme in which an individual replica can sample from states encountered by other replicas at any previous time by way of a global configuration database, enabling the fast propagation of relevant states through the whole ensemble of replicas. This mechanism depends on the knowledge of global thermodynamic functions which are measured during the simulation and not coupled to the heat bath temperatures driving the individual simulations. Therefore, this setup also allows for a continuous adaptation of the temperature set. In this paper, we will review the new scheme and demonstrate its capability. The method is particularly useful for the fast and reliable estimation of the microcanonical temperature T(U) or, equivalently, of the density of states g(U) over a wide range of energies.

physics.comp-ph↗

arXiv2015

Towards an optimal flow: Density-of-states-informed replica-exchange simulations

Replica exchange (RE) is one of the most popular enhanced-sampling simulations technique in use today. Despite widespread successes, RE simulations can sometimes fail to converge in practical amounts of time, e.g., when sampling around phase transitions, or when a few hard-to-find configurations dominate the statistical averages. We introduce a generalized RE scheme, density-of-states-informed RE (g-RE), that addresses some of these challenges. The key feature of our approach is to inform the simulation with readily available, but commonly unused, information on the the density of states of the system as the RE simulation proceeds. This enables two improvements, namely, the introduction of resampling moves that actively move the system towards equilibrium, and the continual adaptation of the optimal temperature set. As a consequence of these two innovations, we show that the configuration flow in temperature space is optimized and that the overall convergence of RE simulations can be dramatically accelerated.

cond-mat.stat-mech↗

arXiv2014

Analysis of Transition State Theory Rates upon Spatial Coarse-Graining

Spatial multiscale methods have established themselves as useful tools for extending the length scales accessible by conventional statics (i.e., zero temperature molecular dynamics). Recently, extensions of these methods, such as the finite-temperature quasicontinuum (hot-QC) or Coarse-Grained Molecular Dynamics (CGMD) methods, have allowed for multiscale molecular dynamics simulations at finite temperature. Here, we assess the quality of the long-time dynamics these methods generate by considering canonical transition rates. Specifically, we analyze the transition state theory (TST) rates in CGMD and compare them to the corresponding TST rate of the fully atomistic system. The ability of such an approach to reliably reproduce the TST rate is verified through a relative error analysis, which is then used to highlight the major contributions to the error and guide the choice of degrees of freedom. Finally, our analytical results are compared with numerical simulations for the case of a 1-D chain.

math.NA↗

arXiv2014

Diffusion and transformation kinetics of small Helium clusters in bulk Tungsten

The production of energy through nuclear fusion poses serious challenges related to the stability and performance of materials in extreme conditions. In particular, the constant bombardment of the walls of the reactor with high doses of He ions is known to lead to deleterous changes in their microstructures. These changes follow from the aggregation of He into bubbles that can grow and blister, potentially leading to the contamination of the plasma, or to the degradation of their mechanical properties. We computationally study the behavior of small clusters of He atoms in W in conditions relevant to fusion energy production. Using a wide range of techniques, we investigate the thermodynamics of the clusters and their kinetics in terms of diffusivity, growth, and breakup, as well as mutation into nano-bubbles. Our study provides the essential ingredients to model the early stages of He exposure leading up to the nucleation of He bubbles.

cond-mat.mtrl-sci↗

arXiv2014

Molecular Dynamics in the Multicanonical Ensemble: Equivalence of Wang-Landau Sampling, Statistical Temperature Molecular Dynamics, and Metadynamics

We show direct formal relationship between the Wang-Landau iteration [PRL 86, 2050 (2001)], metadynamics [PNAS 99, 12562 (2002)] and statistical temperature molecular dynamics [PRL 97, 050601 (2006)], the major Monte Carlo and molecular dynamics work-horses for sampling from a generalized, multicanonical ensemble. We demonstrate that statistical temperature molecular dynamics (which is formally derived from the Wang-Landau method), augmented by the introduction of kernel updates of the statistical temperature, generates dynamics which are indistinguishable from a corresponding metadynamics simulation. We also show that the use of a Gaussian kernel significantly improves the performance of statistical temperature molecular dynamics, highlighting the practical benefits of this improved formal understanding.

cond-mat.stat-mech↗

arXiv2014

Sampling in the multicanonical ensemble: Small He clusters in W

We carry out generalized-ensemble molecular dynamics simulations of the formation of small Helium (He) clusters in bulk Tungsten (W), a process of practical relevance for fusion energy production. We calculate formation free energies of small Helium clusters at temperatures up to the melting point of W, encompassing the whole range of interest for fusion-energy production. From this, parameters like cluster break-up or formation rates can be calculated, which help to refine models of microstructure evolution in He-irradiated Tungsten.

cond-mat.mtrl-sci↗