Fredrik Hedman - Academia.edu (original) (raw)

Papers by Fredrik Hedman

In the last couple of years, many e-Science infrastructures have begun to offer production servic... more In the last couple of years, many e-Science infrastructures have begun to offer production services to e- Scientists with an increasing number of applications that require access to different kinds of computational resources. Within Europe two rather different multi-national e-Science infrastructures evolved over time namely Distributed European Infrastructure for Supercomputing Applications (DEISA) and Enabling Grids for E-SciencE (EGEE). DEISA provides access to massively parallel systems such as supercomputers that are well suited for scientific applications that require many interactions between their typically high numbers of CPUs. EGEE on the other hand provides access to a world-wide Grid of university clusters and PC pools that are well suited for farming applications that require less or even no interactions between the distributed CPUs. While DEISA uses the HPC-driven Grid technology UNICORE, EGEE is based on the gLite Grid middleware optimized for farming jobs. Both have ...

International Journal of Quantum Chemistry, 1993

ABSTRACT An efficient data parallel computational scheme is presented for large-scale molecular d... more ABSTRACT An efficient data parallel computational scheme is presented for large-scale molecular dynamics (MD) simulations of liquids with short-range interactions. The method is based on decomposition of the simulation cell into equally sized subcells, with the shortest side length equal to the cutoff radius. Inter- and intracell interactions are calculated in a coarse-grained manner. A geometric sorting procedure, based on particle distances to subcell boundaries, is used to minimize the overall computations and the nonproductive communications. Using only nearest-neighbor communications, an efficient scheme is developed for periodic updates of the contents of subcells due to the migration of particles. Special “null-particles” are introduced, which act as buffers during the periodic updates and allow for a globally uniform algorithm during the calculations. Communication cost is about 7% of the total CPU time. The method is found to be linearly scalable with the number of particles, performing better as the ratio of virtual to physical processors increases. The MDcode is written in Fortran 90 and implemented on a CM-200. The overall speed is approximately 5.9 μs. per MD step and per particle for 1 million particles and 5.5 μs for 5 million particles. The method should be easily transferred to other massively parallel computers of SIMD and MIMD type. © 1993 John Wiley & Sons, Inc.

Methods for performing large-scale parallel Molecular Dynamics(MD) simulations are investigated. ... more Methods for performing large-scale parallel Molecular Dynamics(MD) simulations are investigated. A perspective on the field of parallel md simulations is given. Hardware and software aspects are characterized and the interplay between the two is briefly discussed. A method for performing ab initio md is described; the method essentially recomputes the interaction potential at each time-step. It has been tested on a system of liquid water by comparing results with other simulation methods and experimental results. Different strategies for parallelization are explored. Furthermore, data-parallel methods for short-range and long-range interactions on massively parallel platforms are described and compared. Next, a method for treating electrostatic interactions in md simulations is developed. It combines the traditional Ewald summation technique with the nonuniform Fast Fourier transform-ENUF for short. The method scales as O(N log N), where N is the number of charges in the system. ENUF has a behavior very similar to Ewald summation and can be easily and efficiently implemented in existing simulation programs. Finally, an outlook is given and some directions for further developments are suggested.

Molecular Simulation, May 1, 1995

ABSTRACT An approach to large-scale data-parallel molecular dynamics for simulations of systems w... more ABSTRACT An approach to large-scale data-parallel molecular dynamics for simulations of systems which include Coulombic interactions is investigated. Short-range interactions are calculated with a method based on the use coarse-grained cells. In this method the simulation cell is decomposed into equally sized subcells, with the shortest side being larger than the cut-off radius of the short-range interaction. Electrostatic interactions are calculated using a data-parallel version of the Ewald summation method. Calculations of long- and short-range interactions are merged by a suitable choice of the size of the subcells and the Ewald sum convergence parameter.The method, which should be easy to implement on most massively parallel computers of SIMD or MIMD type has been developed on a 8K CM200 using CM Fortran, and so far it has been tested on systems of over 2 million particles. Timings of systems up to 2 million particles are presented. The time per time-step for a full CM 200 is estimated to be around 15 seconds for a system of 256 thousand particles.

Http Dx Doi Org 10 1080 08927029808022037, Sep 23, 2006

ABSTRACT A parallel implementation to perform quantum molecular dynamics simulations of liquids o... more ABSTRACT A parallel implementation to perform quantum molecular dynamics simulations of liquids on Born-Oppenheimer surfaces is evaluated. The MD method allows the use of atomic energy gradient forces at an arbitrary level of quantum chemical methodology and electronic state. The computational scheme is very simple to implement, although still expensive to use. The method can easily be developed further for studies which are beyond the capabilities of current classical simulations, e.g., simple chemical reactions or other chemical processes involving excited electronic states or radicals. In a preliminary study, the approach is applied to simulate liquid water (with all internal degrees of freedom included) at the semi-empirical AM 1 molecular orbital (MO), and at the ab initio SCF-MO Hartree-Fock level. Performance, scaling and load balancing properties of the parallel scheme are discussed. Timings of an implementation on an IBM SP-2 are presented. To test the approach, the simulation results are compared with those obtained from corresponding classical simulations of water, from simulations using the Car-Parrinello method and to experimental radial distribution functions.

International Journal of Modern Physics C, 1993

An efficient approach to large scale data parallel short-range molecular dynamics for liquids is ... more An efficient approach to large scale data parallel short-range molecular dynamics for liquids is presented. The method is based on the coarse-grained cell method in which the simulation cell is decomposed into equally sized subcells, with the shortest side being larger than the cut-off radius. To avoid a large fraction of the nonproductive calculations we develop a geometric sorting procedure based on particle distances to subcell boundaries. Due to particle migration, the contents of the subcells need to be updated. This is done with a method based only on nearest-neighbor communications. Special "null-particles" are introduced, which act as buffers during periodic updates and allow for a globally uniform algorithm during the force calculations. The method should be easy to implement on most massively parallel computers of SIMD or MIMD type. We have implemented our code in CM Fortran on an 8K CM200. Communication cost is around 7% of the total cpu time. The overall speed ...

Lecture Notes in Computer Science, 1998

A method to perform embarrassingly parallel ab initio molecular dynamics simulations of liquids o... more A method to perform embarrassingly parallel ab initio molecular dynamics simulations of liquids on Born-Oppenheimer surfaces is described. It uses atomic energy gradient forces at an arbitrary level of quantum chemical methodology. The computational scheme is implemented with an MD program interfaced to a quantum chemistry package. Parallelization is done using the replicated data method. Scaling results for up to

Theoretical and Computational Chemistry, 1999

Lecture Notes in Computer Science, 2008

Authentication interoperability between the UNICORE grid middleware system and other Grid middlew... more Authentication interoperability between the UNICORE grid middleware system and other Grid middleware systems is addressed. An approach to extending the UNICORE authentication model to support a proxy certificate (RFC3280) profile is presented. This optional feature can then be enabled based on site policy. Furthermore, the addition capacitates further advances related to authorization. With interoperability becoming a key issue in many production environments, extending the generality of UNICORE in this way opens up the possibility of direct and general interoperability scenarios.

Lecture Notes in Computer Science, 2002

ABSTRACT We examine the performance of the molecular dynamics code Parallacs and the time-domain ... more ABSTRACT We examine the performance of the molecular dynamics code Parallacs and the time-domain (TD) codes in the General ElectroMagnetic Solvers (GEMS) code suite for a 32 CPUs IBM 1.1 GHz pwr4 node. Parallacs is an easily portable molecular dynamics (MD) benchmark code, which contains most features found in current general-purpose production MD codes. Timings and scaling behavior up to 16 CPUs is reported. The GEMS TD codes give the expected performance. For large problems, they are three to four times faster on a 1.1 GHz pwr4 node than on a 375 MHz pwr3 node.

Applied Mathematics, 2014

Ewald summation method, based on Non-Uniform FFTs (ENUF) to compute the electrostatic interaction... more Ewald summation method, based on Non-Uniform FFTs (ENUF) to compute the electrostatic interactions and forces, is implemented in two different particle simulation schemes to model molecular and soft matter, in classical all-atom Molecular Dynamics and in Dissipative Particle Dynamics for coarse-grained particles. The method combines the traditional Ewald method with a non-uniform fast Fourier transform library (NFFT), making it highly efficient. It scales linearly with the number of particles as () log  N N , while being both robust and accurate. It conserves both energy and the momentum to float point accuracy. As demonstrated here, it is straightforward to implement the method in existing computer simulation codes to treat the electrostatic interactions either between point-charges or charge distributions. It should be an attractive alternative to mesh-based Ewald methods.

Third IEEE International Conference on e-Science and Grid Computing (e-Science 2007), 2007

... Svan van de Berghe, Vivian, Li, David Snelling Fujitsu Laboratories of Europe Hayes Hayes, Mi... more ... Svan van de Berghe, Vivian, Li, David Snelling Fujitsu Laboratories of Europe Hayes Hayes, Middlesex UB4 8FE, UK Katerina Stamou, Zeeshan Ali Shah, Fredrik Hedman Center for Parallel Computers Royal Institute of Technology SE-100 44 Stockholm, Sweden Abstract ...

Lecture Notes in Computer Science, 2009

This paper evaluates the performance of the emerging OGF standard OGSA-Basic Execution Service (B... more This paper evaluates the performance of the emerging OGF standard OGSA-Basic Execution Service (BES) on three fundamentally different Grid middleware platforms: UNICORE 5/6, Globus Toolkit 4 and gLite. The particular focus within this paper is on the OGSA-BES implementation of UNICORE 6. A comparison is made with baseline measurements, for UNICORE 6 and Globus Toolkit 4, using the legacy job submission interfaces. Our results show that the BES components are comparable in performance to existing legacy interfaces. We also have a strong indication that other factors, attributable to the supporting infrastructure, have a bigger impact on performance than BES components.

We sketch an O(N logN) approach, recently proposed by us, that is based on the traditional Ewald ... more We sketch an O(N logN) approach, recently proposed by us, that is based on the traditional Ewald summation technique. It can can be used to calculate the electrostatic energies and forces in molecular computer simulations and appears to conserve both energy and momen- tum. The method is straightforward to implement in existing simulation programs. We report here some results from optimizing our implemen- tation, where we utilize widely available libraries, and demonstrate the speed of our approach as well as the expected computational complexity.

Molecular Simulation, 1998

Molecular Simulation, 1995