Parallel Implementation of the Density Matrix Renormalization Group Method Achieving a Quarter petaFLOPS Performance on a Single DGX-H100 GPU Node
Not Available
37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
SEARCH · Engineering Papers
Results for “parallel methods”
Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.
Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.
Parallel solution of partial differential equations by extrapolation methods
We have found, in the ROGE algorithm, an extrapolation process which is robust, effective and practically simple to implement. It removes the difficulty of needing to make a precise estimate of the over-relaxation parameter for Successive Over-Relaxation (SOR) type methods.
97 MATHEMATICS AND COMPUTING↗
Performance analysis of three dimensional integral equation computations on a massively parallel computer
The purpose of this study is to investigate the performance of the integral equation computations using numerical source field-panel method in a massively parallel processing (MPP) environment. A comparative study of computational performance of the MPP CM-5 computer and conventional Cray-YMP supercomputer for a three-dimensional flow problem is made. A serial FORTRAN code is converted into a parallel CM-FORTRAN code. Some performance results are obtained on CM-5 with 32, 62, 128 nodes along with those on Cray-YMP with a single processor. The comparison of the performance indicates that the parallel CM-FORTRAN code near or out-performs the equivalent serial FORTRAN code for some cases.
Logan, Terry G.↗
Contingency Analysis Based on Partitioned and Parallel Holomorphic Embedding
In the steady-state contingency analysis, the traditional Newton-Raphson method suffers from non-convergence issues when solving post-outage power flow problems, which hinders the integrity and accuracy of security assessment. In this paper, we propose a novel robust contingency analysis approach based on holomorphic embedding (HE). Here, the HE-based simulator provides theoretical convergence guarantee, which is desirable because it avoids the influence of numerical issues and provides a credible security assessment conclusion. In addition, based on the multi-area characteristics of real-world power systems, a partitioned HE (PHE) method is proposed with an interfacebased partitioning of HE formulation. The PHE method does not undermine the numerical robustness of HE and significantly reduces the computation burden in large-scale contingency analysis. The PHE method is further enhanced by parallel or distributed computation to become parallel PHE (P2HE). Tests on a 458-bus system, a synthetic 419-bus system and a large-scale 21447-bus system demonstrate the advantages of the proposed methods in robustness and efficiency.
42 ENGINEERING↗
Turbulent flow between a rotating disk and a parallel wall
A scalar effective viscosity method is used to calculate the incompressible flow in the gap between two infinite, parallel smooth disks, one of them rotating and the other stationary. The flow is calculated for two values of the gap Reynolds number omega(s squared)/nu, namely 296 and 2852, where omega is the disk angular velocity, s the gap width, and nu the kinematic viscosity of the fluid; and over a range of radial Reynolds number omega(r squared)/nu from 0 to 10 million, where r is the radius from the axis of rotation. A numerical shooting technique is used to match the two boundary-layer flows arising on the disk and wall. Gradual transition is assumed for radial Reynolds numbers from 160,000 to 250,000. The results are in good agreement with Daily and Nece's measurements of velocity profiles and disk friction drag for enclosed rotating disks, and at sufficiently large radius the local skin friction approaches that of turbulent Couette flow.
Cooper, P.↗
A parallel implementation of a multisensor feature-based range-estimation method
There are many proposed vision based methods to perform obstacle detection and avoidance for autonomous or semi-autonomous vehicles. All methods, however, will require very high processing rates to achieve real time performance. A system capable of supporting autonomous helicopter navigation will need to extract obstacle information from imagery at rates varying from ten frames per second to thirty or more frames per second depending on the vehicle speed. Such a system will need to sustain billions of operations per second. To reach such high processing rates using current technology, a parallel implementation of the obstacle detection/ranging method is required. This paper describes an efficient and flexible parallel implementation of a multisensor feature-based range-estimation algorithm, targeted for helicopter flight, realized on both a distributed-memory and shared-memory parallel computer.
Suorsa, Raymond E.↗
A parallel implementation of a multisensor feature-based range-estimation method
There are many proposed vision based methods to perform obstacle detection and avoidance for autonomous or semi-autonomous vehicles. All methods, however, will require very high processing rates to achieve real time performance. A system capable of supporting autonomous helicopter navigation will need to extract obstacle information from imagery at rates varying from ten frames per second to thirty or more frames per second depending on the vehicle speed. Such a system will need to sustain billions of operations per second. To reach such high processing rates using current technology, a parallel implementation of the obstacle detection/ranging method is required. This paper describes an efficient and flexible parallel implementation of a multisensor feature-based range-estimation algorithm, targeted for helicopter flight, realized on both a distributed-memory and shared-memory parallel computer.
Suorsa, Raymond E.↗
High Performance Parallel Computational Nanotechnology
At a recent press conference, NASA Administrator Dan Goldin encouraged NASA Ames Research Center to take a lead role in promoting research and development of advanced, high-performance computer technology, including nanotechnology. Manufacturers of leading-edge microprocessors currently perform large-scale simulations in the design and verification of semiconductor devices and microprocessors. Recently, the need for this intensive simulation and modeling analysis has greatly increased, due in part to the ever-increasing complexity of these devices, as well as the lessons of experiences such as the Pentium fiasco. Simulation, modeling, testing, and validation will be even more important for designing molecular computers because of the complex specification of millions of atoms, thousands of assembly steps, as well as the simulation and modeling needed to ensure reliable, robust and efficient fabrication of the molecular devices. The software for this capacity does not exist today, but it can be extrapolated from the software currently used in molecular modeling for other applications: semi-empirical methods, ab initio methods, self-consistent field methods, Hartree-Fock methods, molecular mechanics; and simulation methods for diamondoid structures. In as much as it seems clear that the application of such methods in nanotechnology will require powerful, highly powerful systems, this talk will discuss techniques and issues for performing these types of computations on parallel systems. We will describe system design issues (memory, I/O, mass storage, operating system requirements, special user interface issues, interconnects, bandwidths, and programming languages) involved in parallel methods for scalable classical, semiclassical, quantum, molecular mechanics, and continuum models; molecular nanotechnology computer-aided designs (NanoCAD) techniques; visualization using virtual reality techniques of structural models and assembly sequences; software required to control mini robotic manipulators for positional control; scalable numerical algorithms for reliability, verifications and testability. There appears no fundamental obstacle to simulating molecular compilers and molecular computers on high performance parallel computers, just as the Boeing 777 was simulated on a computer before manufacturing it.
Saini, Subhash↗
Swarm Verification
Reportedly, supercomputer designer Seymour Cray once said that he would sooner use two strong oxen to plow a field than a thousand chickens. Although this is undoubtedly wise when it comes to plowing a field, it is not so clear for other types of tasks. Model checking problems are of the proverbial "search the needle in a haystack" type. Such problems can often be parallelized easily. Alas, none of the usual divide and conquer methods can be used to parallelize the working of a model checker. Given that it has become easier than ever to gain access to large numbers of computers to perform even routine tasks it is becoming more and more attractive to find alternate ways to use these resources to speed up model checking tasks. This paper describes one such method, called swarm verification.
grid computing↗
Static and dynamic characteristics of parallel-grooved seals
Presented is an analytical method to determine static and dynamic characteristics of annular parallel-grooved seals. The governing equations were derived by using the turbulent lubrication theory based on the law of fluid friction. Linear zero- and first-order perturbation equations of the governing equations were developed, and these equations were analytically investigated to obtain the reaction force of the seals. An analysis is presented that calculates the leakage flow rate, the torque loss, and the rotordynamic coefficients for parallel-grooved seals. To demonstrate this analysis, we show the effect of changing number of stages, land and groove width, and inlet swirl on stability of the boiler feed water pump seals. Generally, as the number of stages increased or the grooves became wider, the leakage flow rate and rotor-dynamic coefficients decreased and the torque loss increased.
Iwatsubo, Takuzo↗
Parallel Multi-Omics in High-Risk Subjects for the Identification of Integrated Biomarker Signatures of Type 1 Diabetes
Background: Biomarkers are crucial for detecting early type-1 diabetes (T1D) and preventing significant β-cell loss before the onset of clinical symptoms. Here, we present proof-of-concept studies to demonstrate the potential for identifying integrated biomarker signature(s) of T1D using parallel multi-omics. Methods: Blood from human subjects at high risk for T1D (and healthy controls; n = 4 + 4) was subjected to parallel unlabeled proteomics, metabolomics, lipidomics, and transcriptomics. The integrated dataset was analyzed using Ingenuity Pathway Analysis (IPA) software for disturbances in the at-risk subjects compared to controls. Results: The final quadra-omics dataset contained 2292 proteins, 328 miRNAs, 75 metabolites, and 41 lipids that were detected in all samples without exception. Disease/function enrichment analyses consistently indicated increased activation, proliferation, and migration of CD4 T-lymphocytes and macrophages. Integrated molecular network predictions highlighted central involvement and activation of NF-κB, TGF-β, VEGF, arachidonic acid, and arginase, and inhibition of miRNA Let-7a-5p. IPA-predicted candidate biomarkers were used to construct a putative integrated signature containing several miRNAs and metabolite/lipid features in the at-risk subjects. Conclusions: Preliminary parallel quadra-omics provided a comprehensive picture of disturbances in high-risk T1D subjects and highlighted the potential for identifying associated integrated biomarker signatures. With further development and validation in larger cohorts, parallel multi-omics could ultimately facilitate the classification of T1D progressors from non-progressors.
60 APPLIED LIFE SCIENCES↗
Development and Verification of the Charring Ablating Thermal Protection Implicit System Solver
The development and verification of the Charring Ablating Thermal Protection Implicit System Solver is presented. This work concentrates on the derivation and verification of the stationary grid terms in the equations that govern three-dimensional heat and mass transfer for charring thermal protection systems including pyrolysis gas flow through the porous char layer. The governing equations are discretized according to the Galerkin finite element method with first and second order implicit time integrators. The governing equations are fully coupled and are solved in parallel via Newton's method, while the fully implicit linear system is solved with the Generalized Minimal Residual method. Verification results from exact solutions and the Method of Manufactured Solutions are presented to show spatial and temporal orders of accuracy as well as nonlinear convergence rates.
Amar, Adam J.↗
The hypercluster: A parallel processing test-bed architecture for computational mechanics applications
The development of numerical methods and software tools for parallel processors can be aided through the use of a hardware test-bed. The test-bed architecture must be flexible enough to support investigations into architecture-algorithm interactions. One way to implement a test-bed is to use a commercial parallel processor. Unfortunately, most commercial parallel processors are fixed in their interconnection and/or processor architecture. In this paper, we describe a modified n cube architecture, called the hypercluster, which is a superset of many other processor and interconnection architectures. The hypercluster is intended to support research into parallel processing of computational fluid and structural mechanics problems which may require a number of different architectural configurations. An example of how a typical partial differential equation solution algorithm maps on to the hypercluster is given.
Blech, Richard A.↗
Optimization Plugin Library
The Optimization Plugin library ("op") is a lightweight general optimization solver interface. The primary purpose of op is to simplify the process of integrating different optimization solvers (serial or parallel) with scalable parallel physics engines. By design it has several features that help make this a reality. The core abstraction interface was developed to encompass a large class of optimization problems in an optimizer-agnostic way. This enables us to describe the optimization problem once and then use a variety of supported "op" optimizers with ideally no code-changes. The abstraction interface is made up of lightweight wrappers that make it easy to integrate with existing simulation codes. This makes integration less intrusive and should minimize changes to existing physics codes. The "op" interface includes an assortment of utility methods that help specify parallel communication patterns as well as methods to convert from optimization-specific interfaces to the more general "op" interface. Lastly a dynamic library linking interface is provided to allow for use of proprietary optimization engines without explicit reference in the source code, along with standard shared library interfaces for opensource engines.
Jekel, CharlesF↗
Domain Decomposition By the Advancing-Partition Method
A new method of domain decomposition has been developed for generating unstructured grids in subdomains either sequentially or using multiple computers in parallel. Domain decomposition is a crucial and challenging step for parallel grid generation. Prior methods are generally based on auxiliary, complex, and computationally intensive operations for defining partition interfaces and usually produce grids of lower quality than those generated in single domains. The new technique, referred to as "Advancing Partition," is based on the Advancing-Front method, which partitions a domain as part of the volume mesh generation in a consistent and "natural" way. The benefits of this approach are: 1) the process of domain decomposition is highly automated, 2) partitioning of domain does not compromise the quality of the generated grids, and 3) the computational overhead for domain decomposition is minimal. The new method has been implemented in NASA's unstructured grid generation code VGRID.
Pirzadeh, Shahyar Z.↗
Iterative finite element solver on transputer networks
The parallelism inherent in the Conjugate Gradient method is described. The initial results of a parallel implementation on a network of twelve transputers are discussed. The high efficiencies obtained indicate that significant speedup can be obtained with larger transputer arrays if communication overhead can be kept low. To this end, a method of communication that allows large, dynamically reconfigurable transputer arrays to exchange data in log sub 4 N steps for N processors is suggested.
Danial, Albert↗
NUMERICAL ANALYSIS OF THE FOKKER-PLANCK EQUATION WITH ADIABATIC FOCUSING: ISOTROPIC PITCH-ANGLE SCATTERING
We present numerical solutions to both the standard and modified two-dimensional Fokker-Planck equations with adiabatic focusing and isotropic pitch-angle scattering. With the numerical solution of the particle distribution function, we then discuss the related numerical issues, calculate the parallel diffusion coefficient using several different methods, and compare our numerical solutions for the parallel diffusion coefficient to the analytical forms derived earlier. We find the numerical solution to the diffusion coefficient for both the standard and modified Fokker-Planck equations agrees with that of Shalchi for the mean squared displacement method of computing the diffusion coefficient. However, we also show the numerical solution agrees with that of Litvinenko and Shalchi and Danos when calculating the diffusion coefficient via the velocity correlation function.
71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗
WE-G-BRD-05: Inline Magnetic Fields Enhance Tumor Dose for Small Lung Cancers
Purpose: To report on significant dose enhancement effects caused by magnetic fields aligned parallel to 6MV photon beam radiotherapy of small lung tumors. Findings are applicable to future inline MRI-guided radiotherapy systems. Methods: 9 clinical lung plans were recalculated using Monte Carlo methods and external inline (parallel to the beam direction) magnetic fields of 0.5 T, 1.0 T and 3 T were included. Three plans were 6MV 3D-CRT and six were 6MV IMRT. The GTV’s ranged from 0.8 cc to 73 cc, while the PTV ranged from 1 cc to 180 cc. Results: The inline magnetic field has a moderate impact in lung dose distributions by reducing the lateral scatter of secondary electrons and causing a small local dose increase. Superposition of multiple small beams acts to superimpose the small dose increases and can lead to significant dose enhancements, especially when the GTV is low density. Two plans with very small, low mean density GTV’s (<1 cc, ρ(mean)<0.35g/cc) showed uniform increases of 16% and 23% at 1 T throughout the PTV. Three plans with moderate mean density PTV’s (3–13 cc, ρ(mean)=0.58–0.67 g/cc) showed 6% mean dose enhancement at 1 T in the PTV, however not uniform throughout the GTV/PTV. Replanning would benefit these cases. The remaining 5 plans had large dense GTV’s (∼ 1 g/cc) and so only a minimal (<2%) enhancement was seen. In general the mean dose enhancement at 0.5 T was 60% less than 1 T, while 5–50% higher at 3 T. Conclusions: A paradigm shift in the efficacy of small lung tumor radiotherapy is predicted with future inline MRI-linac systems. This will be achieved by carefully taking advantage of the reduction of lateral electronic disequilibrium withing lung tissue that is induced naturally inside strong inline magnetic fields.
60 APPLIED LIFE SCIENCES↗