Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “supercomputer”

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.

At least 1,621 records · Page 90

Steady-State Calibration of a Diesel Engine in CFD Using a GPU-Based Chemistry Solver

The prospect of analysis-driven pre-calibration of a modern diesel engine is extremely valuable in order to significantly reduce hardware investments and accelerate engine designs compliant with stricter EPA fuel economy regulations. Advanced modeling tools, such as CFD, are often used with the goal of streamlining significant portions of the calibration process. The success of the methodology largely relies on the accuracy of analytical predictions, especially engine-out emissions. However, the effectiveness of CFD simulation tools for in-cylinder engine combustion is often compromised by the complexity, accuracy, and computational overhead of detailed chemical kinetics necessary for combustion calculations. The standard approach has been to use skeletal kinetic mechanisms (~50 species) which consume acceptable computational time but with degraded accuracy.In this work, a comprehensive demonstration and validation of the analytical pre-calibration process is presented for a passenger car diesel engine using CFD simulations with CONVERGE™ and a GPU-based chemical kinetics solver (Zero-RK, developed at Lawrence Livermore National Laboratory) on high performance computing resources to enable the use of detailed kinetic mechanisms. Diesel engine combustion computations have been conducted over 600 operating points spanning in-vehicle speed-load map, using massively parallel ensemble simulation sets on the Titan supercomputer located at the Oak Ridge Leadership Computing Facility. The results with different mesh resolutions have been analyzed to compare differences in combustion and emissions (NO x , Carbon Monoxide CO, Unburned Hydrocarbons UHC, and Smoke) with actual engine measurements. The results show improved agreement in combustion and NO x predictions with a large n-heptane mechanism consisting of 144 species and 900 reactions with refined mesh resolution; however; agreement in CO, UHC and Smoke remain a challenge.

Gao, Jian↗

SU-E-J-15: A Patient-Centered Scheme to Mitigate Impacts of Treatment Setup Error

Purpose: Current Intensity Modulated Radiation Therapy (IMRT) is plan-centered. At each treatment fraction, we position the patient to match the setup in treatment plan. Inaccurate setup can compromise delivered dose distribution, and hence leading to suboptimal treatments. Moreover, current setup approach via couch shift under image guidance can correct translational errors, while rotational and deformation errors are hard to address. To overcome these problems, we propose in this abstract a patient-centered scheme to mitigate impacts of treatment setup errors. Methods: In the patient-centered scheme, we first position the patient on the couch approximately matching the planned-setup. Our Supercomputing Online Replanning Environment (SCORE) is then employed to design an optimal treatment plan based on the daily patient geometry. It hence mitigates the impacts of treatment setup error and reduces the requirements on setup accuracy. We have conducted simulations studies in 10 head-and-neck (HN) patients to investigate the feasibility of this scheme. Rotational and deformation setup errors were simulated. Specifically, 1, 3, 5, 7 degrees of rotations were put on pitch, roll, and yaw directions; deformation errors were simulated by splitting neck movements into four basic types: rotation, lateral bending, flexion and extension. Setup variation ranges are based on observed numbers in previous studies. Dosimetric impacts of our scheme were evaluated on PTVs and OARs in comparison with original plan dose with original geometry and original plan recalculated dose with new setup geometries. Results: With conventional plan-centered approach, setup error could lead to significant PTV D99 decrease (−0.25∼+32.42%) and contralateral-parotid Dmean increase (−35.09∼+42.90%). The patientcentered approach is effective in mitigating such impacts to 0∼+0.20% and −0.03∼+5.01%, respectively. Computation time is <128 s. Conclusion: Patient-centered scheme is proposed to mitigate setup error impacts using replanning. Its superiority in terms of dosimetric impacts and feasibility has been shown through simulation studies on HN cases.

60 APPLIED LIFE SCIENCES↗

SU-E-J-52: Dosimetric Benefit of Adaptive Re-Planning in Lung Cancer Stereotactic Body Radiotherapy (SBRT)

Purpose: To investigate the dosimetric benefit of adaptive re-planning for lung stereotactic body radiotherapy(SBRT). Methods: Five lung cancer patients with SBRT treatment were retrospectively investigated. Our in-house supercomputing online re-planning environment (SCORE) was used to realize the re-planning process. First a deformable image registration was carried out to transfer contours from treatment planning CT to each treatment CBCT. Then an automatic re-planning using original plan DVH guided fluence-map optimization is performed to get a new plan for the up-to-date patient geometry. We compared the re-optimized plan to the original plan projected on the up-to-date patient geometry in critical dosimetric parameters, such as PTV coverage, spinal cord maximum and volumetric constraint dose, esophagus maximum and volumetric constraint dose. Results: The average volume of PTV covered by prescription dose for all patients was improved by 7.56% after the adaptive re-planning. The volume of the spinal cord receiving 14.5Gy and 23Gy (V14.5, V23) decreased by 1.48% and 0.68%, respectively. For the esophagus, the volume receiving 19.5Gy (V19.5) reduced by 1.37%. Meanwhile, the maximum dose dropped off by 2.87% for spinal cord and 4.80% for esophagus. Conclusion: Our experimental results demonstrate that adaptive re-planning for lung SBRT has the potential to minimize the dosimetric effect of inter-fraction deformation and thus improve target coverage while reducing the risk of toxicity to nearby normal tissues.

60 APPLIED LIFE SCIENCES↗

SU-C-16A-03: Direction Modulated Brachytherapy for HDR Treatment of Cervical Cancer

Purpose: To investigate a new Directional Modulated Brachytherapy (DMBT) intra-uterine tandem using various 192-Ir after-loaders. Methods: Dose distributions from the 192-Ir sources were modulated using a 6.3mm diameter tungsten shield (18.0g/cm3). The source moved along 6 longitudinal grooves, each 1.3mm in diameter, evenly spaced along periphery of the shield, The tungsten rod was enclosqed by 0.5mm thick Delrin (1.41g/cc). Monte Carlo N particle (MCNPX) was used to calculate dose distributions. 51million particles were calculated on 504 cores of a supercomputer. Fifteen different patients originally treated with a traditional tandem-and-ovoid applicator, with 5 fractions each, (15 patients X 5 fxs = 75 plans) were re-planned with the DMBT applicator combined with traditional ovoids, on an in-house developed HDR brachytherapy planning platform, which used intensity modulated planning capabilities using a constrained gradient optimization algorithm. For all plans the prescription dose was 6 Gy and they were normalized to match the clinical treated V100. Results: Generally, the DMBT plan quality was a remarkable improvement from conventional T and O plans because of the anisotropic dose distribution of DMBT. The largest difference was to the bladder which had a 0.59±0.87 Gy (8.5±28.7%) reduction in dose. This was because of the the horseshoe shape (U-shape) of the bladder. The dose reduction to rectum and sigmoid were 0.48±0.55 Gy (21.1±27.2%) and 0.10±0.38 Gy (40.6±214.9%), respectively. The D90 to the HRCTV was 6.55±0.96 Gy (conventional T and O) and 6.59±1.06 Gy (DMBT). Conclusion: For image guided adaptive brachytherapy, greater flexibility of radiation intensity is essential and DMBT can be the solution.

60 APPLIED LIFE SCIENCES↗

Implementation of Parallel Adaptive-Krylov Exponential Solvers for Stiff Problems

Recently exponential integrators have been receiving increased attention as a means of solving large stiff systems of ODEs. Preliminary performance analysis demonstrated that exponential integrators hold promise compared to state-of-the-art implicit methods. However, much work remains to be done to understand in detail possible computational advantages these methods may offer in practice. This is particularly true for supercomputer-scale problems as there has been very little work on parallelizing exponential methods. In this paper we describe an implementation of a suite of parallel exponential solvers. Additionally, we present some performance tests on four stiff benchmark problems of a particular adaptive-Krylov exponential propagation iterative Runge--Kutta (EPIRK) method from the suite, and compare efficiency with the Newton--Krylov implicit solver CVODE.

97 MATHEMATICS AND COMPUTING↗

Algebraic multigrid domain and range decomposition (AMG-DD / AMG-RD)*

In modern large-scale supercomputing applications, algebraic multigrid (AMG) is a leading choice for solving matrix equations. However, the high cost of communication relative to that of computation is a concern for the scalability of traditional implementations of AMG on emerging architectures. This paper introduces two new algebraic multilevel algorithms, algebraic multigrid domain decomposition (AMG-DD) and algebraic multigrid range decomposition (AMG-RD), that replace traditional AMG V-cycles with a fully overlapping domain decomposition approach. While the methods introduced here are similar in spirit to the geometric methods developed by Brandt and Diskin [Multigrid solvers on decomposed domains, in Domain Decomposition Methods in Science and Engineering, Contemp. Math. 157, AMS, Providence, RI, 1994, pp. 135--155], Mitchell [Electron. Trans. Numer. Anal., 6 (1997), pp. 224--233], and Bank and Holst [SIAM J. Sci. Comput., 22 (2000), pp. 1411--1443], they differ primarily in that they are purely algebraic: AMG-RD and AMG-DD trade communication for computation by forming global composite “grids” based only on the matrix, not the geometry. (As is the usual AMG convention, “grids” here should be taken only in the algebraic sense, regardless of whether or not it corresponds to any geometry.) Another important distinguishing feature of AMG-RD and AMG-DD is their novel residual communication process that enables effective parallel computation on composite grids, avoiding the all-to-all communication costs of the geometric methods. The main purpose of this paper is to study the potential of these two algebraic methods as possible alternatives to existing AMG approaches for future parallel machines. As a result, this paper develops some theoretical properties of these methods and reports on serial numerical tests of their convergence properties over a spectrum of problem parameters.

97 MATHEMATICS AND COMPUTING↗

Improving Unstructured Mesh Partitions for Multiple Criteria Using Mesh Adjacencies

The scalability of unstructured mesh based applications depends on partitioning methods that quickly balance the computational work while reducing communication costs. Zhou et al. [SIAM J. Sci. Comput., 32 (2010), pp. 3201{3227; J. Supercomput., 59 (2012), pp. 1218{1228] demonstrated the combination of (hyper)graph methods with vertex and element partition improvement for PHASTA CFD scaling to hundreds of thousands of processes. Our work generalizes partition improvement to support balancing combinations of all the mesh entity dimensions (vertices, edges, faces, regions) in partitions with imbalances exceeding 70%. Improvement results are then presented for multiple entity dimensions on up to one million processes on meshes with over 12 billion tetrahedral elements.

97 MATHEMATICS AND COMPUTING↗

Scalable Asynchronous Domain Decomposition Solvers

We discuss how parallel implementations of linear iterative solvers generally alternate between phases of data exchange and phases of local computation. Increasingly large problem sizes and more heterogeneous compute architectures make load balancing and the design of low latency network interconnects that are able to satisfy the communication requirements of linear solvers very challenging tasks. In particular, global communication patterns such as inner products become increasingly limiting at scale. We explore the use of asynchronous communication based on one-sided Message Passing Interface primitives in the context of domain decomposition solvers. In particular, a scalable asynchronous two-level Schwarz method is presented. We discuss practical issues encountered in the development of a scalable solver and show experimental results obtained on a state-of-the-art supercomputer system that illustrate the benefits of asynchronous solvers in load balanced as well as load imbalanced scenarios. Using the novel method, we can observe speedups of up to four times over its classical synchronous equivalent.

97 MATHEMATICS AND COMPUTING↗

A Fast Butterfly-Compressed Hadamard–Babich Integrator for High-Frequency Helmholtz Equations in Inhomogeneous Media with Arbitrary Sources

Here we present a butterfly-compressed representation of the Hadamard-Babich (HB) ansatz for the Green's function of the high-frequency Helmholtz equation in smooth inhomogeneous media. For a computational domain discretized with Nv discretization cells, the proposed algorithm first solves and tabulates the phase and HB coefficients via eikonal and transport equations with observation points and point sources located at the Chebyshev nodes using a set of much coarser computation grids, and then butterfly compresses the resulting HB interactions from all Nv cell centers to each other. The overall CPU time and memory requirement scale as O(Nv log2 Nv) for any bounded two-dimensional (2D) domains with arbitrary excitation sources. A direct extension of this scheme to bounded 3D domains yields an O(Nv4/3) CPU complexity, which can be further reduced to quasi-linear complexities with proposed remedies. The scheme can also efficiently handle scattering problems involving inclusions in inhomogeneous media. Although the current construction of our HB integrator does not accommodate caustics, the resulting HB integrator itself can be applied to certain sources, such as concave-shaped sources, to produce caustic effects. Compared to finite-difference frequency domain methods, the proposed HB integrator is free of numerical dispersion and requires fewer discretization points per wavelength. As a result, it can solve wave propagation problems well beyond the capability of existing solvers. Remarkably, the proposed scheme can accurately model wave propagation in 2D domains with 640 wavelengths per direction and in 3D domains with 54 wavelengths per direction on a state-of-the-art supercomputer at Lawrence Berkeley National Laboratory.

Hadamard--Babich ansatz↗

Augmented Lagrangian Method for Constrained Nuclear Density Functiional Theory

The augmented Lagrangian method (ALM), widely used in quantum chemistry constrained optimization problems, is applied in the context of the nuclear Density Functional Theory (DFT) in the self-consistent constrained Skyrme Hartree-Fock-Bogoliubov (CHFB) variant. The ALM allows precise calculations of multidimensional energy surfaces in the space of collective coordinates that are needed to, e.g., determine fission pathways and saddle points; it improves accuracy of computed derivatives with respect to collective variables that are used to determine collective inertia and is well adapted to supercomputer applications.

Staszczak, A.↗

Online Monitoring System for Performance Fault Detection

To achieve the exaFLOPS performance within a contained power budget, next generation supercomputers will feature hundreds of millions of components operating at low- and near-threshold voltage. As the probability that at least one of these components fails during the execution of an application approaches certainty, it seems unrealistic to expect that any run of a scientific application will not experience some performance faults. We believe that there is need of a new generation of light-weight performance and debugging tools that can be used online even during production runs of parallel applications and that can identify performance anomalies during the application execution. In this work we propose the design and implementation of such a monitoring system.

97 MATHEMATICS AND COMPUTING↗

Parrallel Implementation of Fast Randomized Algorithms for Low Rank Matrix Decomposition

We analyze the parallel performance of randomized interpolative decomposition by de- composing low rank complex-valued Gaussian random matrices larger than 100 GB. We chose a Cray XMT supercomputer as it provides an almost ideal PRAM model permitting quick investigation of parallel algorithms without obfuscation from hardware idiosyncrasies. We obtain that on non-square matrices performance scales almost linearly with runtime about 100 times faster on 128 processors. We also verify that numerically discovered error bounds still hold on matrices two orders of magnitude larger than those previously tested.

Lucas, Andrew J.↗

Algorithms for high-throughput disk-to-disk sorting, In: SC '13 Proceedings of the International Conference on High Performance Computing, Networking, Storage and Analysis

In this paper, we present a new out-of-core sort algorithm, designed for problems that are too large to fit into the aggregate RAM available on modern supercomputers. We analyze the performance including the cost of IO and demonstrate the fastest (to the best of our knowledge) reported throughput using the canonical sortBenchmark on a general-purpose, production HPC resource running Lustre. By clever use of available storage and a formulation of asynchronous data transfer mechanisms, we are able to almost completely hide the computation (sorting) behind the IO latency. This latency hiding enables us to achieve comparable execution times, including the additional temporary IO required, between a large sort problem (5TB) run as a single, in-RAM sort and our out-of-core approach using 1/10th the amount of RAM. In our largest run, sorting 100TB of records using 1792 hosts, we achieved an end-to-end throughput of 1.24TB/min using our general-purpose sorter, improving on the current Daytona record holder by 65%.

Computer Science↗

Taking a quantum leap in time to solution for simulations of high-Tc superconductors

We present a new quantum cluster algorithm to simulate models of high-Tc superconductors. This algorithm extends current methods with continuous lattice self-energies, thereby removing artificial long-range correlations. This cures the fermionic sign problem in the underlying quantum Monte Carlo solver for large clusters and realistic values of the Coulomb interaction in the entire temperature range of interest. We find that the new algorithm improves time-to-solution by nine orders of magnitude compared to current, state of the art quantum cluster simulations. An efficient implementation is given, which ports to multi-core as well as hybrid CPU-GPU systems. Running on 18,600 nodes on ORNL's Titan supercomputer enables us to compute a converged value of Tc/t = 0.053 ± 0.0014 for a 28 site cluster in the 2D Hubbard model with U/t = 7 at 10% hole doping. Typical simulations on Titan sustain between 9.2 and 15.4 petaflops (double precision measured over full run), depending on configuration and parameters used.

42 ENGINEERING↗

The Practical Obstacles of Data Transfer: Why researchers still love scp

The importance of computing facilities is heralded every six months with the announcement of the new Top500 list, showcasing the world s fastest supercomputers. Unfortu- nately, with great computing capability does not come great long-term data storage capacity, which often means users must move their data to their local site archive, to remote sites where they may be doing future computation or anal- ysis, or back to their home institution, else face the dreaded data purge that most HPC centers employ to keep utiliza- tion of large parallel filesystems low to manage performance and capacity. At HPC centers, data transfer is crucial to the scientific workflow and will increase in importance as computing systems grow in size. The Energy Sciences Net- work (ESnet) recently launched its fifth generation network, a 100 Gbps high-performance, unclassified national network connecting more than 40 DOE research sites to support scientific research and collaboration. Despite the tenfold increase in bandwidth to DOE research sites amenable to multiple data transfer streams and high throughput, in prac- tice, researchers often under-utilize the network and resort to painfully-slow single stream transfer methods such as scp to avoid the complexity of using multiple stream tools such as GridFTP and bbcp, and contend with frustration from the lack of consistency of available tools between sites. In this study we survey and assess the data transfer methods pro- vided at several DOE supported computing facilities, includ- ing both leadership-computing facilities, connected through ESnet. We present observed transfer rates, suggested opti- mizations, and discuss the obstacles the tools must overcome to receive wide-spread adoption over scp.

Nam, Hai Ah↗

Discrete Event Execution with One-Sided and Two-Sided GVT Algorithms on 216,000 Processor Cores

Global virtual time (GVT) computation is a key determinant of the efficiency and runtime dynamics of parallel discrete event simulations (PDES), especially on large-scale parallel platforms. Here, three execution modes of a generalized GVT computation algorithm are studied on high-performance parallel computing systems: (1) a synchronous GVT algorithm that affords ease of implementation, (2) an asynchronous GVT algorithm that is more complex to implement but can relieve blocking latencies, and (3) a variant of the asynchronous GVT algorithm to exploit one-sided communication in extant supercomputing platforms. Performance results are presented of implementations of these algorithms on up to 216,000 cores of a Cray XT5 system, exercised on a range of parameters: optimistic and conservative synchronization, fine- to medium-grained event computation, synthetic and non-synthetic applications, and different lookahead values. Performance of up to 54 billion events executed per second is registered. Detailed PDES-specific runtime metrics are presented to further the understanding of tightly-coupled discrete event dynamics on massively parallel platforms.

Perumalla, Kalyan S↗

Fault Tolerance for OpenSHMEM. In: PGAS '14 Proceedings of the 8th International Conference on Partitioned Global Address Space Programming Models. Article No. 23

On today's supercomputing systems, faults are becoming a norm rather than an exception. Given the complexity required for achieving expected scalability and performance on future systems, this situation is expected to become worse. The systems are expected to function in a nearly constant presence of faults. To be productive on these systems, programming models will require both hardware and software to be resilient to faults. With the growing importance of PGAS programming model and OpenSHMEM, as a part of HPC software stack, a lack of a fault tolerance model may become a liability for its users. Towards this end, in this paper, we discuss the viability of using checkpoint/restart as a fault-tolerance method for OpenSHMEM, propose a selective checkpoint/restart fault-tolerance model, and discuss challenges associated with implementing the proposed model.

Hao, Pengfei↗

Thread-level parallelization and optimization of NWChem for the Intel MIC architecture

In the multicore era it was possible to exploit the increase in on-chip parallelism by simply running multiple MPI processes per chip. Unfortunately, manycore processors' greatly increased thread- and data-level parallelism coupled with a reduced memory capacity demand an altogether different approach. In this paper we explore augmenting two NWChem modules, triples correction of the CCSD(T) and Fock matrix construction, with OpenMP in order that they might run efficiently on future manycore architectures. As the next NERSC machine will be a self-hosted Intel MIC (Xeon Phi) based supercomputer, we leverage an existing MIC testbed at NERSC to evaluate our experiments. In order to proxy the fact that future MIC machines will not have a host processor, we run all of our experiments in native mode. We found that while straightforward application of OpenMP to the deep loop nests associated with the tensor contractions of CCSD(T) was sufficient in attaining high performance, significant e ort was required to safely and efeciently thread the TEXAS integral package when constructing the Fock matrix. Ultimately, our new MPI+OpenMP hybrid implementations attain up to 65× better performance for the triples part of the CCSD(T) due in large part to the fact that the limited on-card memory limits the existing MPI implementation to a single process per card. Additionally, we obtain up to 1.6× better performance on Fock matrix constructions when compared with the best MPI implementations running multiple processes per card.

CCSD(T)↗