Efficiency of ab-initio total energy calculations for metals and semiconductors using a plane-wave basis set

Three parallel algorithms for classical molecular dynamics are presented. The first assigns each processor a fixed subset of atoms; the second assigns each a fixed subset of inter-atomic forces to compute; the third assigns each a fixed spatial region. The algorithms are suitable for molecular dynamics models which can be difficult to parallelize efficiently—those with short-range forces where the neighbors of each atom change rapidly. They can be implemented on any distributed-memory parallel machine which allows for message-passing of data between independently executing processors. The algorithms are tested on a standard Lennard-Jones benchmark problem for system sizes ranging from 500 to 100,000,000 atoms on several parallel supercomputers--the nCUBE 2, Intel iPSC/860 and Paragon, and Cray T3D. Comparing the results to the fastest reported vectorized Cray Y-MP and C90 algorithm shows that the current generation of parallel machines is competitive with conventional vector supercomputers even for small problems. For large problems, the spatial algorithm achieves parallel efficiencies of 90% and a 1840-node Intel Paragon performs up to 165 faster than a single Cray C9O processor. Trade-offs between the three algorithms and guidelines for adapting them to more complex molecular dynamics simulations are also discussed.

Fast parallel algorithms for short-range molecular dynamics

QUANTUM ESPRESSO is an integrated suite of computer codes for electronic-structure calculations and materials modeling, based on density-functional theory, plane waves, and pseudopotentials (norm-conserving, ultrasoft, and projector-augmented wave). The acronym ESPRESSO stands for opEn Source Package for Research in Electronic Structure, Simulation, and Optimization. It is freely available to researchers around the world under the terms of the GNU General Public License. QUANTUM ESPRESSO builds upon newly-restructured electronic-structure codes that have been developed and tested by some of the original authors of novel electronic-structure algorithms and applied in the last twenty years by some of the leading materials modeling groups worldwide. Innovation and efficiency are still its main focus, with special attention paid to massively parallel architectures, and a great effort being devoted to user friendliness. QUANTUM ESPRESSO is evolving towards a distribution of independent and interoperable codes in the spirit of an open-source project, where researchers active in the field of electronic-structure calculations are encouraged to participate in the project by contributing their own codes or by implementing their own ideas into existing codes.

/pdf/quantum-espresso-a-modular-and-open-source-software-project-4t6vlupbeh.pdf

QUANTUM ESPRESSO: a modular and open-source software project for quantum simulations of materials

抗原变异可使得多种致病微生物易于逃避宿主免疫应答。表达在感染红细胞表面的恶性疟原虫红细胞表面蛋白1（PfPMP1）与感染红细胞、内皮细胞、树突状细胞以及胎盘的单个或多个受体作用，在黏附及免疫逃避中起关键的作用。每个单倍体基因组var基因家族编码约60种成员，通过启动转录不同的var基因变异体为抗原变异提供了分子基础。

疟原虫var基因转换速率变化导致抗原变异[英]／Paul H, Robert P, Christodoulou Z, et al//Proc Natl Acad Sci U S A

We present ab initio quantum-mechanical molecular-dynamics simulations of the liquid-metal--amorphous-semiconductor transition in Ge. Our simulations are based on (a) finite-temperature density-functional theory of the one-electron states, (b) exact energy minimization and hence calculation of the exact Hellmann-Feynman forces after each molecular-dynamics step using preconditioned conjugate-gradient techniques, (c) accurate nonlocal pseudopotentials, and (d) Nos\'e dynamics for generating a canonical ensemble. This method gives perfect control of the adiabaticity of the electron-ion ensemble and allows us to perform simulations over more than 30 ps. The computer-generated ensemble describes the structural, dynamic, and electronic properties of liquid and amorphous Ge in very good agreement with experiment. The simulation allows us to study in detail the changes in the structure-property relationship through the metal-semiconductor transition. We report a detailed analysis of the local structural properties and their changes induced by an annealing process. The geometrical, bonding, and spectral properties of defects in the disordered tetrahedral network are investigated and compared with experiment.

Ab initio molecular-dynamics simulation of the liquid-metal-amorphous-semiconductor transition in germanium.

We have generalized the nonequilibrium Green’s functions Keldysh formalism to study the effect of interfacial disorder on the average spin transfer torque, 〈T∥〉, in magnetic tunnel junctions (MTJs). We find a sinusoidal angular behavior of the average 〈T∥〉 as in ideal MTJs. We demonstrate for the first time that the general expression of the bias behavior of the average 〈T∥〉 in terms of the interplay of average spin current densities in collinear configurations is valid even in the presence of disorder. This explains the strong enhancement and sign reversal of 〈T∥〉 in the positive bias region, due to the disorder-induced resonance states at interface which selectively assist the transmission of right-coming electrons.

/pdf/effect-of-disorder-on-spin-transfer-torque-in-magnetic-1228r57sap.pdf

Effect of disorder on spin-transfer torque in magnetic tunnel junctions

Publisher’s Note: Density Functional Theory of the Electrical Conductivity of Molecular Devices [Phys. Rev. Lett. 94 , 146803 (2005)]

When Roberto Car and Michele Parrinello started collaborating in the early 1980s, electronic structure calculations and molecular dynamics (MD) simulations of atomic configurations were virtually isolated from each other. But Car and Parrinello combined their individual expertise to create a method that bridged the gap between the two communities and changed the approach to materials modelling in a radical way1. The Car–Parrinello ab initio molecular dynamics method was conceived and developed within the theoretical physics community in Trieste, in a period in which SISSA (Scuola Internazionale Superiore di Studi Avanzati) was still a very young institute and was practically overlapping with the well-established International Centre of Theoretical Physics (ICTP). The presence of many bright physicists and the continuous flow of international visitors made the atmosphere highly stimulating. Discussions went on virtually without interruption, even during the occasional swims around the shores of the Miramare Castle (pictured), only a few steps away from the ICTP building. Car’s background in density functional theory (DFT) and extreme enthusiasm was perfectly complemented by Parrinello’s knowledge of statistical mechanics and molecular dynamics, in addition to his reflective attitude, resulting in an ideal combination of personality and scientific expertise. The paper reporting the Car–Parrinello method1 was published in November 1985, and on the occasion of its twenty-fifth anniversary we explore the origin of the work and its long-lasting impact. What emerges from the Interviews with the two scientists2,3 is that the main motivation for their efforts were the limitations suffered by both DFT and classical MD, which effectively restricted their application to the realistic simulation of condensed matter at finite temperature to specific cases. Pure DFT was mainly applicable to the electronic structures of ordered and homogenous systems. On the other hand, the forces between atoms used in MD did not take into account the fact that the electronic potentials varied with the atomic movement during the progress of a simulation. By using DFT to calculate the potential ‘felt’ by atoms and letting such potential evolve with each step of the simulation, the Car–Parrinello method allowed a much wider range of disordered and therefore more realistic materials systems to be studied. This was beautifully demonstrated by applying the method to amorphous silicon, which led to results on both the atomic structure and the electronic properties that were in very good agreement with experimental observations4. More generally, the ability to follow the evolution of the electronic potential has allowed important studies of chemical reactions occurring, for example, in liquids and large biomolecules. In his Commentary5, besides recalling the fundamental concepts behind the Car–Parrinello method, Jürgen Hafner describes the highly inspirational role that the approach had in stimulating the creation of a series of first-principles computational codes that are still used widely today. A special role for these codes is their application to phenomena that are difficult to study experimentally, such as those occurring at high pressure or during complex and fast chemical reactions. An aspect of these codes that is not always appreciated is their importance in industrial applications. The CASTEP code6, conceived and developed by Mike Payne and collaborators at the University of Cambridge, is a perfect example. Since its commercialization, it has been the highest source of revenue for the university in the physical sciences. The code is the fundamental part of the software package ‘Materials Studio’, developed and commercialized by Accelrys, with the aim of providing a powerful tool to generic users with no specific knowledge of code writing. According to Gerhard Goldbeck-Wood, director of product marketing for Materials Studio, the software package is used by chemical companies that work in fields such as catalysis or high-k dielectrics, to name just a few. Most importantly an essential aim of these types of code is their use in feasibility tests, which has substantial economical benefits. Unlike the discovery of a new molecule or the observation of a new phenomenon, it is difficult to appreciate the success that a computational technique is likely to have when it is first introduced. After a quarter of a century however, it is clear that the Car–Parrinello method has been ground-breaking within the field of computational materials science and has had an enormous impact on fundamental science and applications in a wide range of fields, from solid-state materials physics to chemistry and biology. Still, the approach was originally conceived by two scientists motivated solely by their passion for science and their desire to understand nature, with no specific agenda with regard to the possible commercial value of their work. This should send a strong message to decision makers in charge of funding research. ❐

/pdf/a-model-approach-to-modelling-1olrpbo17c.pdf

A model approach to modelling

The study of the structural and electronic properties of microclusters is a field of growing interest. Ab-initio molecular dynamics has provided a new and important tool for the theoretical approach to these questions. Here we present some very recent results on small semiconductor aggregates with special reference to calculations of equilibrium shapes and temperature effects. Results of simulations on alkali-metal microclusters are briefly mentioned.

AB-Initio Molecular Dynamics Studies of Microclusters

The energy functional of a novel electronic structure theory, OP-NSOFT, has as variables the natural spin orbitals (NSO) of the trial function and their joint occupation probabilities in the search for the ground state energy. When occupancy is restricted to the spin-paired NSOs of DOCI, the resulting theory, OP-NSOFT-0, scales as M3, with M the size of the one-electron basis set. Accurate results were obtained for small molecules, particularly near dissociation where single reference theories like DFT are inaccurate. The homogeneous electron liquid (HEL) could serve as a test bed of OP-NSOFT for condensed systems, but OP-NSOFT-0 reduces to the Hartree–Fock approximation for the HEL. Cooper pairing is introduced instead, both singlet pairing, OP-NOFT-Cs, and fully polarized triplet pairing, OP-NSOFT-Ct. The former yields 1/3 of the diffusion-Monte-Carlo correlation energy, the latter 1/2 to 1/3 with decreasing electron density for rs values between 1 and 10. Both yield the discontinuity in the single-particle occupation number required by the Luttinger theorem. Two-state joint occupation probabilities illustrate the importance of electron–electron small-angle scattering in establishing electron correlation in the unpolarized HEL.

Roberto Car

Papers

Effect of disorder on spin-transfer torque in magnetic tunnel junctions

Publisher’s Note: Density Functional Theory of the Electrical Conductivity of Molecular Devices [Phys. Rev. Lett. 94 , 146803 (2005)]

A model approach to modelling

AB-Initio Molecular Dynamics Studies of Microclusters

Occupation probabilities as variables in electronic structure theory: cooper pairing, OP-NSOFT-Cs,t, and the homogeneous electron liquid