Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “parallelization”

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 1,063 records · Page 59
All sourcesNASA NTRSDOE OSTI
DOE OSTI2025

A provably stable numerical method for the anisotropic diffusion equation in confined magnetic fields

We present a novel numerical method for solving the anisotropic diffusion equation in magnetic fields confined to a periodic box which is accurate and provably stable. We derive energy estimates of the solution of the continuous initial boundary value problem. A discrete formulation is presented using operator splitting in time with the summation by parts finite difference approximation of spatial derivatives for the perpendicular diffusion operator. Weak penalty procedures are derived for implementing both boundary conditions and parallel diffusion operator obtained by field line tracing. We prove that the fully-discrete approximation is unconditionally stable. Discrete energy estimates are shown to match the continuous energy estimate given the correct choice of penalty parameters. A nonlinear penalty parameter is shown to provide an effective method for tuning the parallel diffusion penalty and significantly minimises rounding errors. Several numerical experiments, using manufactured solutions, the “NIMROD benchmark” problem and a single island problem, are presented to verify numerical accuracy, convergence, and asymptotic preserving properties of the method. Finally, we present a magnetic field with chaotic regions and islands and show the contours of the anisotropic diffusion equation reproduce key features in the field.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

STORM: Scrape-off layer turbulence in tokamak fusion reactors

The scrape-off layer of a tokamak fusion reactor carries the plasma exhaust from the hot core plasma to the material surfaces of the reactor vessel. The heat loads imposed by the exhaust are a critical limit on the performance of fusion power plants. Turbulent transport of the plasma regulates the width of the scrape-off layer plasma and must be modelled to understand the intensity of these heat loads. STORM is a plasma turbulence code capable of simulating three dimensional turbulence across the full scrape-off layer of a tokamak fusion reactor, using a drift reduced, collisional fluid model. STORM uses mostly finite difference schemes, with a staggered grid in the direction parallel to the magnetic field. We describe the model, geometry and initialisation options used by STORM, as well as the numerical methods, which are implemented using the BOUT++ plasma simulation framework. BOUT++ has been enhanced alongside the development of STORM, providing better support for staggered grid methods. We summarise these enhancements, including a detailed explanation of the parallel derivative methods, which underwent a major update for version 4 of BOUT++.

BOUT++↗
DOE OSTI2024

A GPU accelerated mixed-precision Smoothed Particle Hydrodynamics framework with cell-based relative coordinates

Smoothed Particle Hydrodynamics (SPH) is essential for modeling complex large-deformation problems across various applications, requiring significant computational power. A major portion of SPH computation time is dedicated to the Nearest Neighboring Particle Search (NNPS) process. While advanced NNPS algorithms have been developed to enhance SPH efficiency, the potential efficiency gains from modern computation hardware remain underexplored. Here, this study investigates the impact of GPU parallel architecture, low-precision computing on GPUs, and GPU memory management on NNPS efficiency. Our approach employs a GPU-accelerated mixed-precision SPH framework, utilizing low precision float-point 16 (FP16) for NNPS while maintaining high precision for other components. To ensure FP16 accuracy in NNPS, we introduce a Relative Coordinated-based Link List (RCLL) algorithm, storing FP16 relative coordinates of particles within background cells. Our testing results show three significant speedup rounds for CPU-based NNPS algorithms. The first comes from parallel GPU computations, with up to a 1000x efficiency gain. The second is achieved through low-precision GPU computing, where the proposed FP16-based RCLL algorithm offers a 1.5x efficiency improvement over the FP64-based approach on GPUs. By optimizing GPU memory bandwidth utilization, the efficiency of the FP16 RCLL algorithm can be further boosted by 2.7x, as demonstrated in an example with 1 million particles. Our code is released at https://github.com/pnnl/lpNNPS4SPH.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Bayesian Optimization for Anything (BOA): An open-source framework for accessible, user-friendly Bayesian optimization

We introduce Bayesian Optimization for Anything (BOA), a high-level Bayesian Optimization (BO) framework and model wrapping toolkit, which presents a novel approach to simplifying BO, with the goal of making it more accessible and user-friendly, particularly for those with limited expertise in the field. BOA addresses common barriers in implementing BO, focusing on ease of use, reducing the need for deep domain knowledge, and cutting down on extensive coding requirements. A notable feature of BOA is its language-agnostic architecture, which facilitates broader application in various fields and to a wider audience. We showcase BOA's application through three examples: a high-dimensional optimization with parameters of the SWAT+ watershed model, a highly parallelized optimization of this intrinsically non-parallel model, and a multi-objective optimization of the FETCH Tree-Crown Hydrodynamics model. Furthermore, these test cases illustrate BOA's effectiveness in addressing complex optimization challenges in diverse scenarios.

54 ENVIRONMENTAL SCIENCES↗
DOE OSTI2020

Advanced diagnostics to evaluate heterogeneity in lithium-ion battery modules

Battery packs for electric and stationary applications experience varied operating conditions, including abuse—e.g., fast charging, overcharging, thermal, vibration, shock, etc.—throughout their lifetimes. Innovative diagnostic tools and algorithms that go beyond single cells and deal with modules and packs are essential for early detection of off-normal issues. High-resolution tools with known detection limits are key to developing appropriate mitigation strategies. With the advent of rapid impedance spectroscopy that can generate a broadband impedance spectrum in ~10 sec, the case for impedance-based diagnostics that can be readily aligned with other methods, such as incremental capacity or dQ.dV -1 , has become promising. This study used the aforementioned diagnostic methods to identify realistic in-vehicle battery abnormalities (e.g., localized self-discharge and non-uniform aging), in series (up to 10S) and parallel (4P) strings, using 16 Ah graphite/NMC cells. The impedance-based diagnostic is found to be sensitive to the string size and state. Depending on the type of abnormality, detection frequency varied. The dQ.dV -1 method showed the potential to detect long-term aging-related heterogeneity in modules. In general, both the impedance and dQ.dV -1 methods were able to detect series strings’ abnormalities, but struggled to find those issues within parallel modules.

25 ENERGY STORAGE↗
DOE OSTI2024

Toward a high-fidelity tritium transport modeling for retention and permeation experiments

We report Tritium Migration Analysis Program version 8 (TMAP8), the latest version of TMAP, was developed within the framework of the Multiphysics Object-Oriented Simulation Environment (MOOSE). Created at Idaho National Laboratory (INL), MOOSE is an open-source, dimension-agnostic, fully coupled, and fully implicit multiphysics platform featuring massively parallel computation capabilities. Using TMAP8, tritium transport in a divertor monoblock was analyzed to elucidate the effects of pulsed operation (up to fifty 1,600 s plasma discharge and cool-down cycles) on the tritium in-vessel inventory source term and ex-vessel release term (i.e., tritium retention and permeation) for safety analysis. With its built-in Message Passing Interface capability, TMAP8 can, in under 2 h, simulate tritium transport in three different layered materials (i.e., tungsten, copper, and copper-chromium-zirconium alloy) in 2D geometry, using a single device/computer with 10 cores. The MOOSE-based TMAP8 code can leverage other MOOSE tools developed under the Nuclear Energy Advanced Modeling and Simulation program to perform tritium and thermal transport in complex geometries and multiphysics environments. And via its massively parallel computation, MOOSE will enable the fusion pilot plant designers to conduct high-fidelity multiphysics modeling for the design of the divertor and blanket systems as well as for the safety analysis.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2024

An exploration of online-simulation-driven portfolio scheduling in Workflow Management Systems

Workflow Management Systems used to automate the execution of scientific workflow applications on parallel and distributed computing platforms must make scheduling decisions at runtime. A large number of workflow scheduling algorithms have been proposed in the literature, but often these algorithms are evaluated based on simplifying assumptions that may not hold in practice. Furthermore, published algorithm evaluation and/or comparison results are necessarily only for a subset of all possible scenarios, and thus may not include scenarios relevant to particular use-cases. Consequently, it is difficult for Workflow Management Systems (WMSs) developers to decide which scheduling algorithm should be implemented. To obviate this difficulty, one possible approach is to implement a portfolio of scheduling algorithms and select the most effective algorithm at runtime. One method for performing this selection is to run an online simulation for each algorithm in the portfolio. The algorithm that leads to the best performance, in simulation, is selected for future use. The above simulation-driven portfolio scheduling (SDPS) approach has been proposed in a few parallel and distributed computing contexts. The main objective of this work is to evaluate the feasibility and potential merit of SDPS if implemented in WMSs. Here we perform this evaluation using simulated WMS executions, where the simulations are instantiated from real-world platform and workflow configurations. Our main finding is that SDPS is on par with or outperforms an approach in which a single algorithm is used, where this algorithm is the one that performs best on average across all our experimental scenarios. Furthermore, we find that SDPS remains an attractive proposition even in the presence of high levels of simulation error and for simulators with relatively low levels of sophistication. In many of our experimental scenarios we find that mitigating simulation error at runtime can further improve performance. Finally, we show that simulation overhead can be made sufficiently low for SDPS to be feasible in practice.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Review of hybrid HVDC systems combining line communicated converter and voltage source converter

Hybrid HVDC system, which consists of the advantages of line commutated converter (LCC) and voltage source converter (VSC), is an emerging power transmission system. This paper presents a review of four LCC-VSC hybrid HVDC topologies. The first topology is the pole-hybrid HVDC system, in which the LCC and VSC form the positive- and negative-pole respectively. The second one is the terminal-hybrid HVDC system; in this topology one terminal adopts LCC and the other terminal adopts VSC. The series converter-hybrid HVDC system is the third topology wherein each terminal is formed by LCC and VSC in series. The fourth hybrid topology under consideration is a parallel converter-hybrid HVDC system with LCC and VSC connected in parallel in each terminal. Here, the main contribution of this paper is a comprehensive analysis and comparison of the four mentioned hybrid topologies in terms of PQ operating zone, power flow reversal method, and DC fault ride-through strategy.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2026

Influence of rolling reduction and annealing on recrystallization and grain structure in Ta-2.5W alloys

The microstructural evolution of wrought Tantalum - 2.5 wt% Tungsten (Ta-2.5W) alloys during thermomechanical processing is critical for optimizing their mechanical reliability in demanding applications such as aerospace, chemical processing, and nuclear technology. Despite the widespread use of Ta-W alloys, a comprehensive understanding of how rolling reduction, annealing temperature, and elemental inhomogeneity interact to determine recrystallization behavior and grain refinement remains incomplete. Here, in this study, we systematically investigate the effects of cold rolling and subsequent annealing on the microstructure of Ta-2.5W, with particular attention to grain orientation, stored energy, and elemental banding. Our results demonstrate that higher rolling reduction rates lower the onset and completion temperatures for recrystallization, resulting in finer and more homogeneous grain structures. Electron backscatter diffraction (EBSD) analysis reveals that grains with a 〈111〉 parallel to the plate normal possess higher stored energy and nucleate recrystallization more readily than grains with a 〈001〉 parallel to the plate normal. Elemental mapping shows that tungsten inhomogeneity leads to localized bands of accelerated recrystallization and hardness variation. These findings provide new insights into the mechanisms of microstructure refinement in Ta-2.5W alloys, offering guidance for tailoring processing routes to achieve superior performance in demanding engineering environments.

Annealing↗
DOE OSTI2020

Real-time optimization of multi-cell industrial evaporative cooling towers using machine learning and particle swarm optimization

Existing electrical generating stations must operate with greater flexibility due to increasing renewable energy penetration on the electrical grid, and many coal-fired power stations have transitioned away from baseload operation to load-following operation to aid in grid stability. In cases where multiple independently controlled cooling tower cells are used in parallel for the cooling purposes of such stations, there is an opportunity to increase plant efficiency through data-driven optimization across their full load ranges. This work presents a novel application of real-time optimization using machine learning and particle swarm optimization on a multi-cell induced-draft cooling tower servicing a coal-fired power station under variable load. This is the first work to demonstrate simultaneous optimization of a multi-cell cooling tower, in addition to using machine learning for closed-loop control on a cooling tower. A novel control configuration is presented that ensures original control logic is not adversely affected and that the overall plant process is not disrupted using only existing hardware and operational data. To verify this methodology, the 12 independent cooling tower cells are simulated in parallel using historic operating data to demonstrate the effectiveness of real-time optimization compared to current practice. An artificial neural network is trained to predict overall cooling tower power consumption using only operational data and ambient conditions with an R2 value of greater than 0.96. The real-time optimization using particle swarm yields 6.7% annual energy usage savings compared to current practices, although the extent of the real-time savings varies greatly with both plant load and environmental conditions. This is particularly significant for a variable load situation because frequent ramping typically results in reduced overall efficiency. Furthermore, this proposed AI-based solution presents an opportunity to improve the overall heat rate of a load-following coal-fired power plant without the need to perform extensive first-principles modeling or add additional hardware to the cooling tower, resulting in more resources conserved and less overall emissions per unit of electricity generated.

42 ENGINEERING↗
DOE OSTI2022

A weighted state redistribution algorithm for embedded boundary grids

State redistribution is an algorithm that stabilizes cut cells for embedded boundary grid methods. This work extends the earlier algorithm in several important ways. First, state redistribution is extended to three spatial dimensions. Second, we discuss several algorithmic changes and improvements motivated by the more complicated cut cell geometries that can occur in higher dimensions. In particular, we introduce a weighted version with less dissipation in an easily generalizable framework. Third, we demonstrate that state redistribution can also stabilize a solution update that includes both advective and diffusive contributions. Notably, the stabilization algorithm is shown to be effective for incompressible as well as compressible reacting flows. Finally, we discuss the implementation of the algorithm for several exascale-ready simulation codes based on AMReX, demonstrating ease of use in combination with domain decomposition, hybrid parallelism and complex physics.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Scalable Implicit Solvers with Dynamic Mesh Adaptation for a Relativistic Drift-Kinetic Fokker–Planck–Boltzmann Model

In this work we consider a relativistic drift-kinetic model for runaway electrons along with a Fokker–Planck operator for small-angle Coulomb collisions, a radiation damping operator, and a secondary knock-on (Boltzmann) collision source. Here, we develop a new scalable fully implicit solver utilizing finite volume and conservative finite difference schemes and dynamic mesh adaptivity. A new data management framework in the PETSc library based on the p4est library is developed to enable simulations with dynamic adaptive mesh refinement (AMR), distributed memory parallelization, and dynamic load balancing of computational work. This framework and the runaway electron solver building on the framework are able to dynamically capture both bulk Maxwellian at the low-energy region and a runaway tail at the high-energy region. To effectively capture features via the AMR algorithm, a new AMR indicator prediction strategy is proposed that is performed alongside the implicit time evolution of the solution. This strategy is complemented by the introduction of computationally cheap feature-based AMR indicators that are analyzed theoretically. Numerical results quantify the advantages of the prediction strategy in better capturing features compared with nonpredictive strategies; and we demonstrate trade-offs regarding computational costs. The robustness with respect to model parameters, algorithmic scalability, and parallel scalability are demonstrated through several benchmark problems including manufactured solutions and solutions of different physics models. We focus on demonstrating the advantages of using implicit time stepping and AMR for runaway electron simulations.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗
DOE OSTI2025

Implementation of a Mesh refinement algorithm into the quasi-static PIC code QuickPIC

Plasma-based acceleration (PBA) has emerged as a promising candidate for the accelerator technology used to build a future linear collider and/or an advanced light source. In PBA, a trailing or witness particle beam is accelerated in the plasma wave wakefield (WF) created by a laser or particle beam driver. The WF is often nonlinear and involves the crossing of plasma particle trajectories in real space and thus particle-in-cell methods are used. The distance over which the drive beam evolves is several orders of magnitude larger than the wake wavelength. This large disparity in length scales is amenable to the quasi-static approach. Three-dimensional (3D), quasi-static (QS), particle-in-cell (PIC) codes, e.g., QuickPIC, have been shown to provide high fidelity simulation capability with 2-4 orders of magnitude speedup over 3D fully explicit PIC codes. In PBA, the witness beam needs to be matched to the focusing forces of the WF to reduce the emittance growth. In some linear collider designs, the matched spot size of the witness beam can be 2 to 3 orders of magnitude smaller than the spot size (and wavelength) of the wakefield. Such an additional disparity in length scales is ideal for mesh refinement where the WF within the witness beam is described on a finer mesh than the rest of the WF. A mesh refinement scheme is described that has been implemented into the 3D QS PIC code, QuickPIC. Very fine (high) resolution is used in a small spatial region that includes the witness beam and progressively coarser resolutions in the rest of the simulation domain. A fast multigrid Poisson solver has been implemented for the field solve on the refined meshes and a Fast Fourier Transform (FFT) based Poisson solver is used for the coarse mesh. The code has been parallelized with both MPI and OpenMP, and the parallel scalability has also been improved by using pipelining. A preliminary adaptive mesh refinement technique is described to optimize the computational time for simulations with an evolving witness beam size. Several test problems are used to verify that the mesh refinement algorithm provides accurate results. Additionally, the results are benchmarked against highly resolved simulations exhibiting near-azimuthal symmetry, performed using QPAD—a novel hybrid QS PIC code that uses a PIC description in the coordinates (r, ct – z) and a gridless description in the azimuthal angle, Φ.

Linear collider↗
DOE OSTI2025

Laser-induced slip casting as an additive manufacturing approach for silicon carbide

Here, this work presents processing silicon carbide (SiC) with the laser-induced slip casting (LIS) additive manufacturing (AM). SiC was stabilized in water with polyethyleneimine (PEI) dispersant, and SiC slurries were made with rheology for LIS printing. High-density ceramic parts were printed, followed by single-step binder burnout and sintering. The printed parts achieved 93–95 % of theoretical density. X-ray computed tomography (XCT) revealed a small distribution of flaws exceeding 100 microns. The mechanical properties were measured in both parallel and perpendicular to the printing layers, and the orientation with layers perpendicular to the bending moment resulted in higher strength compared to the parallel direction. Porosity resulting from processing and large inclusions of boron carbide (B4C) were the root cause of failure in the measured samples. Despite these defects through this effort, this new approach demonstrates promise for green forming of SiC with densities greater than 95 % theoretical and tensile strengths above 250 MPa.

Additive Manufacturing↗
DOE OSTI2020

Kink band and shear band localization in anisotropic perfectly plastic solids

Shear-driven strain localization has been observed in a wide variety of materials and may take the form of shear bands or kink bands. Based on observations of kink bands in plastically anisotropic metallic nanolaminates and single crystal metals, here we posit that, for the specific case of isochoric deformation, the kinematics of kink band formation are indistinguishable from those of plane strain shear band formation. The only distinction between shear bands and kink bands in these systems would then be that kink bands ‘lock up’ at a particular value of material rotation while shear bands may progress to arbitrarily high strains. In order to investigate whether strong material anisotropy is sufficient to arrest shear localization at a geometry that matches the classic kink band geometry, we model the development of a band of simple shear within an anisotropic perfectly plastic material. The resulting analytical model provides the stress state needed to maintain the kinematics of simple shear as a function of material anisotropy, deformation band orientation, and shear strain (or equivalently, material rotation). It is found that plastic anisotropy can promote either kink band or shear band formation depending on the loading orientation. When the deviatoric stress is positive parallel to the plane of anisotropy, shear localization may progress without bound and a shear band is produced. When the deviatoric stress is negative parallel to the plane of anisotropy, shear localization is arrested after a certain material rotation, resulting in a kink band. Examination of the requisite applied stress state during kink band formation provides an explanation for the experimentally-observed ‘lock up’ geometry. Solutions for the band boundary inclination angle are obtained and used to provide bounds on permissible band angles for both shear bands and kink bands.

36 MATERIALS SCIENCE↗
DOE OSTI2022

An investigation of the failure modes in U-10Mo monolithic fuel irradiated to high burnup

Here, this study investigated possible failure mode sequences in U-10Mo monolithic fuel irradiated to very-high-burnup. The U-Mo fuel plate used in this study was characterized using scanning electron microscopy (SEM), wavelength dispersive spectroscopy (WDS), and image analysis techniques to investigate how microstructural features may initiate crack formation and propagation through the microstructure, potentially leading to blistering. Distinctly large fission gas pores (FGPs) were observed to preferentially align 5-15µm away from the U-10Mo/Zr interface, parallel to the Zr diffusion barrier. Preferential growth, alignment, and interconnection of FGP could initiate blistering in the fuel plate by creating a fission gas channel, parallel to the U-Mo/Zr interface. FGP alignment in the fuel phase near the U-Mo/Zr interface is believed to be one of the precursors to crack and Type 2 blister formation in monolithic U-Mo fuels. The chemical maps revealed differences in fission product behavior near the U-Mo/Zr interface such that Nd remains in the fuel matrix at the FGP sites. On the other hand, other fission products like Xe and Cs can get trapped in FGPs by Nd in the fuel phase and/or diffuse toward the Zr diffusion barrier.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2024

Synchrotron micro-computed tomography analysis of neutron-irradiated U-Mo fuel

The three-dimensional (3D) microstructure of neutron-irradiated uranium-10 wt.% molybdenum (U-10Mo) fuel with a burn-up of 9.8 × 10 21 fissions/cm 3 was characterized using a novel, multi-modal synchrotron micro-computed tomography approach combining propagation-based phase-contrast enhanced and absorption contrast techniques. The porosity development, porosity interconnectedness, swelling, composition, local thickness of the zirconium (Zr) diffusion barrier, and the influence of the fuel–cladding interaction on the local composition and pore morphology, were uniquely determined in 3D. Two cuboids were produced using a focused ion beam-scanning electron microscope at the Zr diffusion barrier–fuel interface and in the bulk fuel. The bulk fuel sample swelled by 53.3 [+9.7/−3.1]%, while the fuel near the Zr–fuel interface swelled by 63.3 [+14.7/−7.2]%. The average local thickness of the Zr diffusion barrier decreased by 53 %, compared to the expected pre-irradiated thickness. Four pore morphology regions were identified initiating parallel to the fuel–Zr interaction region: (1) an interaction layer of suppressed porosity, (2) a layer of elongated and interconnected porosity, (3) a transition zone of low porosity, and (4) a layer of unoriented porosity representative of the bulk fuel behavior. The increase in porosity near the diffusion barrier corresponded to a higher U concentration compared to that in the bulk fuel. The interconnected porosity in the fuel near the diffusion barrier was extensive and oriented parallel to the diffusion barrier, while the bulk fuel had more compact and isolated pore networks. The interaction layer, despite having suppressed porosity, was nearly 100 wt.% U. Porosity suppression at the diffusion barrier corresponds to the expected reduction in radiation-driven diffusion of Xe at the interface despite the anticipated increase in fission product nucleation originating from a higher U concentration. In conclusion, the novel 3D insights of the porosity, swelling, and compositional variations characterized herein can improve the fidelity of fuel performance codes for proliferation-resistant fuels for research and test reactors.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗
DOE OSTI2020

Optimizing the hypre solver for manycore and GPU architectures

The solution of large-scale combustion problems with codes such as Uintah on modern computer architectures requires the use of multithreading and GPUs to achieve performance. Uintah uses a low-Mach number approximation that requires iteratively solving a large system of linear equations. The Hypre iterative solver has solved such systems in a scalable way for Uintah, but the use of OpenMP with Hypre leads to at least slowdown due to OpenMP overheads. The proposed solution uses the MPI Endpoints within Hypre, where each team of threads acts as a different MPI rank. This approach minimizes OpenMP synchronization overhead and performs as fast or (up to 1.44) faster than Hypre's MPI-only version, and allows the rest of Uintah to be optimized using OpenMP. The profiling of the GPU version of Hypre shows the bottleneck to be the launch overhead of thousands of micro-kernels. The GPU performance was improved by fusing these micro-kernels and was further optimized by using Cuda-aware MPI, resulting in an overall speedup of 1.16—1.44 compared to the baseline GPU implementation. The above optimization strategies were published in the International Conference on Computational Science 2020 [1]. This work extends the previously published research by carrying out the second phase of communication-centered optimizations in Hypre to improve its scalability on large-scale supercomputers. Additionally, this includes an efficient non-blocking inter-thread communication scheme, communication-reducing patch assignment, and expression of logical communication parallelism to a new version of the MPICH library that utilizes the underlying network parallelism [2]. The above optimizations avoid communication bottlenecks previously observed during strong scaling and improve performance by up to 2 on 256 nodes of Intel Knight's Landing processor.

97 MATHEMATICS AND COMPUTING↗