Computational science & discovery最新文献

{"title":"Cheap contouring of costly functions: the Pilot Approximation Trajectory algorithm","authors":"J. Huttunen, P. Stark","doi":"10.1088/1749-4699/5/1/015006","DOIUrl":"https://doi.org/10.1088/1749-4699/5/1/015006","url":null,"abstract":"The Pilot Approximation Trajectory (PAT) contour algorithm can find the contour of a function accurately when it is not practical to evaluate the function on a grid dense enough to use a standard contour algorithm, for instance, when evaluating the function involves conducting a physical experiment or a computationally intensive simulation. PAT relies on an inexpensive pilot approximation to the function, such as interpolating from a sparse grid of inexact values, or solving a partial differential equation (PDE) numerically using a coarse discretization. For each level of interest, the location and ?trajectory? of an approximate contour of this pilot function are used to decide where to evaluate the original function to find points on its contour. Those points are joined by line segments to form the PAT approximation of the contour of the original function.Approximating a contour numerically amounts to estimating a lower level set of the function, the set of points on which the function does not exceed the contour level. The area of the symmetric difference between the true lower level set and the estimated lower level set measures the accuracy of the contour. PAT measures its own accuracy by finding an upper confidence bound for this area.In examples, PAT can estimate a contour more accurately than standard algorithms, using far fewer function evaluations than standard algorithms require. We illustrate PAT by constructing a confidence set for viscosity and thermal conductivity of a flowing gas from simulated noisy temperature measurements, a problem in which each evaluation of the function to be contoured requires solving a different set of coupled nonlinear PDEs.","PeriodicalId":89345,"journal":{"name":"Computational science & discovery","volume":"5 1","pages":"015006"},"PeriodicalIF":0.0,"publicationDate":"2012-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1088/1749-4699/5/1/015006","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"60596691","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Numerical aspects of drift kinetic turbulence: ill-posedness, regularization and a priori estimates of sub-grid-scale terms","authors":"R. Samtaney","doi":"10.1088/1749-4699/5/1/014004","DOIUrl":"https://doi.org/10.1088/1749-4699/5/1/014004","url":null,"abstract":"We present a numerical method based on an Eulerian approach to solve the Vlasov?Poisson system for 4D drift kinetic turbulence. Our numerical approach uses a conservative formulation with high-order (fourth and higher) evaluation of the numerical fluxes coupled with a fourth-order accurate Poisson solver. The fluxes are computed using a low-dissipation high-order upwind differencing method or a tuned high-resolution finite difference method with no numerical dissipation. Numerical results are presented for the case of imposed ion temperature and density gradients. Different forms of controlled regularization to achieve a well-posed system are used to obtain convergent resolved simulations. The regularization of the equations is achieved by means of a simple collisional model, by inclusion of an ad-hoc hyperviscosity or artificial viscosity term or by implicit dissipation in upwind schemes. Comparisons between the various methods and regularizations are presented. We apply a filtering formalism to the Vlasov equation and derive sub-grid-scale (SGS) terms analogous to the Reynolds stress terms in hydrodynamic turbulence. We present a priori quantifications of these SGS terms in resolved simulations of drift-kinetic turbulence by applying a sharp filter.","PeriodicalId":89345,"journal":{"name":"Computational science & discovery","volume":"5 1","pages":"014004"},"PeriodicalIF":0.0,"publicationDate":"2012-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1088/1749-4699/5/1/014004","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"60596619","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Implementation of Accelerated Molecular Dynamics in NAMD.","authors":"Yi Wang, Christopher B Harrison, Klaus Schulten, J Andrew McCammon","doi":"10.1088/1749-4699/4/1/015002","DOIUrl":"10.1088/1749-4699/4/1/015002","url":null,"abstract":"<p><p>Accelerated molecular dynamics (aMD) is an enhanced-sampling method that improves the conformational space sampling by reducing energy barriers separating different states of a system. Here we present the implementation of aMD in the parallel simulation program NAMD. We show that aMD simulations performed with NAMD have only a small overhead compared with classical MD simulations. Through example applications to the alanine dipeptide, we discuss the choice of acceleration parameters, the interpretation of aMD results, as well as the advantages and limitations of the aMD method.</p>","PeriodicalId":89345,"journal":{"name":"Computational science & discovery","volume":"4 1","pages":""},"PeriodicalIF":0.0,"publicationDate":"2011-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://www.ncbi.nlm.nih.gov/pmc/articles/PMC3115733/pdf/nihms283990.pdf","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"30251522","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Finding regions of interest on toroidal meshes","authors":"Kesheng Wu, R. Sinha, Chad Jones, S. Ethier, S. Klasky, K. Ma, A. Shoshani, M. Winslett","doi":"10.1088/1749-4699/4/1/015003","DOIUrl":"https://doi.org/10.1088/1749-4699/4/1/015003","url":null,"abstract":"Fusion promises to provide clean and safe energy, and a considerable amount of research effort is underway to turn this aspiration intoreality. This work focuses on a building block for analyzing data produced from the simulation of microturbulence in magnetic confinementfusion devices: the task of efficiently extracting regions of interest. Like many other simulations where a large amount of data are produced,the careful study of ``interesting'' parts of the data is critical to gain understanding. In this paper, we present an efficient approach forfinding these regions of interest. Our approach takes full advantage of the underlying mesh structure in magnetic coordinates to produce acompact representation of the mesh points inside the regions and an efficient connected component labeling algorithm for constructingregions from points. This approach scales linearly with the surface area of the regions of interest instead of the volume as shown with bothcomputational complexity analysis and experimental measurements. Furthermore, this new approach is 100s of times faster than a recentlypublished method based on Cartesian coordinates.","PeriodicalId":89345,"journal":{"name":"Computational science & discovery","volume":"4 1","pages":"015003"},"PeriodicalIF":0.0,"publicationDate":"2011-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1088/1749-4699/4/1/015003","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"60596496","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"ESTEST: a framework for the validation and verification of electronic structure codes","authors":"G. Yuan, F. Gygi","doi":"10.1088/1749-4699/3/1/015004","DOIUrl":"https://doi.org/10.1088/1749-4699/3/1/015004","url":null,"abstract":"We present a framework for the verification and validation (V&V) of electronic structure simulation software. Electronic structure computations involve numerous parameters and approximations that determine their accuracy and reliability. As a large number of simulation data coming from several electronic structure codes are becoming available, the associated V&V process is becoming increasingly complex. We introduce ESTEST as a framework for facilitating the verification and validation of electronic structure computations. ESTEST software enables the V&V, comparison and sharing of simulation data by constructing a unified representation of code outputs, populating and organizing a query database with these representations and interfacing the data through a web service that offers ways to search, view, compare, visualize and post-process the data. We present examples of V&V as well as comparison and analysis from our implementation, and justify the details of each of the innovative features of this software. The present implementation supports electronic structure codes such as Qbox, Quantum Espresso, ABINIT, and the Exciting code. An online demonstration is available at http://estest.ucdavis.edu.","PeriodicalId":89345,"journal":{"name":"Computational science & discovery","volume":"3 1","pages":"015004"},"PeriodicalIF":0.0,"publicationDate":"2010-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1088/1749-4699/3/1/015004","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"60596453","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Validation of frequency extraction calculations from time-domain simulations of accelerator cavities","authors":"T. Austin, J. Cary, S. Ovtchinnikov, G. Werner, L. Bellantoni","doi":"10.1088/1749-4699/4/1/015004","DOIUrl":"https://doi.org/10.1088/1749-4699/4/1/015004","url":null,"abstract":"The frequency extraction algorithm (Werner and Cary 2008 J. Comp. Phys. 227 5200) that enables a simple finite-difference time-domain algorithm to be transformed into an efficient eigenmode solver is applied to a realistic accelerator cavity modeled with embedded boundaries. Previously, the frequency extraction method was shown to be capable of distinguishing M degenerate modes by running M different simulations and to permit mode extraction with minimal post-processing effort that requires solving only a small eigenvalue problem. Realistic calculations for an accelerator cavity are presented in this paper to validate the method for realistic modeling scenarios and to illustrate the complexities of the computational validation process. The approach is able to extract frequencies that differ by corrected experimentally measured frequencies by about 1 part in 104, which is accounted for (in largest part) by machining errors. The extraction of frequencies and modes from accelerator cavities provides engineers and physicists with an understanding of potential cavity performance as it depends on the shape without incurring manufacture and measurement costs.","PeriodicalId":89345,"journal":{"name":"Computational science & discovery","volume":"4 1","pages":"015004"},"PeriodicalIF":0.0,"publicationDate":"2010-05-11","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1088/1749-4699/4/1/015004","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"60596507","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"An anisotropic preconditioning for the Wilson fermion matrix on the lattice","authors":"B. Joó, R. Edwards, M. Peardon","doi":"10.1088/1749-4699/3/1/015001","DOIUrl":"https://doi.org/10.1088/1749-4699/3/1/015001","url":null,"abstract":"A preconditioning for the Wilson fermion matrix on the lattice is defined, which is particularly suited to the case when the temporal lattice spacing is much smaller than the spatial one. Details on the implementation of the scheme are given. The method is tested in numerical studies of quantum chromodynamics on anisotropic lattices.","PeriodicalId":89345,"journal":{"name":"Computational science & discovery","volume":"3 1","pages":"015001"},"PeriodicalIF":0.0,"publicationDate":"2009-10-06","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1088/1749-4699/3/1/015001","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"60596443","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Welcome to the journal","authors":"A. Mezzacappa","doi":"10.1088/1749-4699/1/1/010101","DOIUrl":"https://doi.org/10.1088/1749-4699/1/1/010101","url":null,"abstract":"It is with the greatest pleasure and with much excitement that we announce the launch of Computational Science and Discovery (CSD). When CSD was conceived, our hope was to provide a unique venue for computational science that crosses disciplines and focuses on scientific discovery through computation. We wanted to provide a venue for the publication of advancements in computational science, applied mathematics, and computer science that make such scientific discovery possible. Our first collection of articles mirrors this vision and hope. Here we find breakthrough science achieved through computation in areas as diverse as astrophysics, fusion energy science, materials science, nuclear physics, and quantum chromodynamics. Our intention is that achievements made in one area will interest, educate, and inspire computational science teams in other areas. And while the science reported in this volume and anticipated in future volumes spans a very broad range, there are common underlying challenges all computational science teams face. We plan to publish progress on meeting these challenges and foster communication across disciplines in order that the progress achieved in one field may benefit collaborative efforts in others. Moreover, collaborative efforts between scientists, mathematicians, and computer scientists—essential to progress in computational science—are often difficult to place according to traditional classifications and measures of progress in these fields. For example, how do we classify efforts to tune physics-based preconditioners for certain radiation transport applications on massively parallel computing platforms? And where do we place the development of effective, efficient, and automated custom scientific workflows that include data management, analysis, and visualization of large and complex simulation data in a particular scientific domain? Within the pages of CSD, we look forward to the publication of breakthrough science and the computational methodologies on which this endeavour will always be founded, coauthored by multidisciplinary teams of scientists, mathematicians, and computer scientists. In so doing we hope to provide, in one venue, the information necessary for the corroboration of such breakthroughs, and documentation of the multidisciplinary collaborative efforts and their resultant approaches that make these advances possible. Please join us in making this exciting new venture a success. We hope you will find CSD an ideal venue for the publication of your team's next exciting results.","PeriodicalId":89345,"journal":{"name":"Computational science & discovery","volume":"242 1","pages":"010101"},"PeriodicalIF":0.0,"publicationDate":"2009-03-03","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1088/1749-4699/1/1/010101","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"60596465","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"High performance parallel computing of flows in complex geometries: I. Methods","authors":"N. Gourdain, L. Gicquel, M. Montagnac, O. Vermorel, M. Gazaix, G. Staffelbach, Marta García, J. Boussuge, T. Poinsot","doi":"10.1088/1749-4699/2/1/015003","DOIUrl":"https://doi.org/10.1088/1749-4699/2/1/015003","url":null,"abstract":"Efficient numerical tools coupled with high-performance computers, have become a key element of the design process in the fields of energy supply and transportation. However flow phenomena that occur in complex systems such as gas turbines and aircrafts are still not understood mainly because of the models that are needed. In fact, most computational fluid dynamics (CFD) predictions as found today in industry focus on a reduced or simplified version of the real system (such as a periodic sector) and are usually solved with a steady- state assumption. This paper shows how to overcome such barriers and how such a new challenge can be addressed by developing flow solvers running on high-end computing platforms, using thousands of computing cores. Parallel strategies used by modern flow solvers are discussed with particular emphases on mesh-partitioning, load balancing and communication. Two examples are used to illustrate these concepts: a multi-block structured code and an unstructured code. Parallel computing strategies used with both flow solvers are detailed and compared. This comparison indicates that mesh-partitioning and load balancing are more straightforward with unstructured grids than with multi-block structured meshes. However, the mesh-partitioning stage can be challenging for unstructured grids, mainly due to memory limitations of the newly developed massively parallel architectures. Finally, detailed investigations show that the impact of mesh-partitioning on the numerical CFD solutions, due to rounding errors and block splitting, may be of importance and should be accurately addressed before qualifying massively parallel CFD tools for a routine industrial use.","PeriodicalId":89345,"journal":{"name":"Computational science & discovery","volume":"2 1","pages":"015003"},"PeriodicalIF":0.0,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1088/1749-4699/2/1/015003","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"60596378","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

{"title":"Virtual materials design using databases of calculated materials properties","authors":"Ture R. Munter, D. D. Landis, F. Abild-Pedersen, Glenn Jones, Shengguang Wang, T. Bligaard","doi":"10.1088/1749-4699/2/1/015006","DOIUrl":"https://doi.org/10.1088/1749-4699/2/1/015006","url":null,"abstract":"Materials design is most commonly carried out by experimental trial and error techniques. Current trends indicate that the increased complexity of newly developed materials, the exponential growth of the available computational power, and the constantly improving algorithms for solving the electronic structure problem, will continue to increase the relative importance of computational methods in the design of new materials. One possibility for utilizing electronic structure theory in the design of new materials is to create large databases of materials properties, and subsequently screen these for new potential candidates satisfying given design criteria. We utilize a database of more than 81 000 electronic structure calculations. This alloy database is combined with other published materials properties to form the foundation of a virtual materials design framework (VMDF). The VMDF offers a flexible collection of materials databases, filters, analysis tools and visualization methods, which are particularly useful in the design of new functional materials and surface structures. The applicability of the VMDF is illustrated by two examples. One is the determination of the Pareto-optimal set of binary alloy methanation catalysts with respect to catalytic activity and alloy stability; the other is the search for new alloy mercury absorbers.","PeriodicalId":89345,"journal":{"name":"Computational science & discovery","volume":"2 1","pages":"015006"},"PeriodicalIF":0.0,"publicationDate":"2009-01-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://sci-hub-pdf.com/10.1088/1749-4699/2/1/015006","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"60596398","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}