Search NASA⌕ Search

SEARCH · Search NASA

Results for “Scalable computation”

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 685 records · Page 38

Optimization of a Solver for Computational Materials and Structures Problems on NVIDIA Volta and AMD Instinct GPUs

The Scalable Implementation of Finite Elements by NASA (ScIFEN) is a software package developed to solve complex computational materials and structures problems using the finite element method (FEM). In this paper, we describe optimization techniques to speed up the linear solver computation that occurs within the ScIFEN application. We consider GPUs from two different vendors, NVIDIA and AMD as our target platforms for optimization and highlight differences in performance and optimization techniques. The NVIDIA GPU Volta V100 is used in the Summit system deployed at Oak Ridge National Laboratory, and the new exascale system, Frontier, will be using AMD Radeon Instinct GPU. We evaluated the performance of various optimization techniques on test matrices, ranging in size from100K to 4M, that are representative of ScIFEN applications. The linear solver computation is memory-bound on both GPUs. Our experiments show that on the NVIDIA GPU we obtained up to79%of the theoretical peak bandwidth, while the AMD GPU achieved 59%. Overall, the NVIDIA V100 GPU outperforms the AMD MI 25 GPU1. We observed an overall speedup of up to37X on an NVIDIA V100 compared to an Intel Skylake 12-coremachine. The solver for a 4M degree of freedom system took under 2.5 seconds.

Mohammad Zubair↗

Scalable Multilevel Uncertainty Quantification Concepts for Extreme-Scale Multiscale Problems

Our work in this project is aimed at making fundamental advances in multilevel and multiscale methods for problems with uncertainties related to flows in porous media. The main thrust of this research is to develop a systematic multiscale, multilevel parallel approaches for problems with uncertainties and use them in the uncertainty quantification. A key emphasis is on problems without an apparent scale separation in space, time, and uncertainty spaces. Multiscale solution methods are currently under active investigation. However, the develop- ment of multiscale methods for problems with uncertainties where the scales and uncertainties are tightly coupled is a research area that is less explored. During this project, we have made some significant contributions to the developments of multiscale methods, the uncertainty quantification for multiscale methods, multiscale model learning, uncertainty quantification in inverse problems, and applications.

97 MATHEMATICS AND COMPUTING↗

Porting hypre to heterogeneous computer architectures: Strategies and experiences

We report that linear systems are occurring in many applications, and solving them can take a large amount of the total simulation time. The high performance library hypre provides a variety of interfaces and linear solvers, including various multigrid methods, that have achieved good scalability on a variety of homogeneous parallel computer architectures. Heterogeneous architectures with nodes that have both CPUs and accelerators provide new challenges, since they require more fine-grained parallelism and reduced data movement between different memories on a single node as well as across nodes. We will discuss our experiences and strategies to port hypre to heterogeneous computers with accelerators, including the design of a new memory model, the use of abstractions, the BoxLoop macros in the structured and semi-structured interfaces, and the restructuring of algebraic multigrid (AMG) into modular components. We present numerical experiments comparing CPU and GPU performance for several test problems.

97 MATHEMATICS AND COMPUTING↗

Massively scalable Kerr comb-driven silicon photonic link

Abstract The growth of computing needs for artificial intelligence and machine learning is critically challenging data communications in today’s data-centre systems. Data movement, dominated by energy costs and limited ‘chip-escape’ bandwidth densities, is perhaps the singular factor determining the scalability of future systems. Using light to send information between compute nodes in such systems can dramatically increase the available bandwidth while simultaneously decreasing energy consumption. Through wavelength-division multiplexing with chip-based microresonator Kerr frequency combs, independent information channels can be encoded onto many distinct colours of light in the same optical fibre for massively parallel data transmission with low energy. Although previous high-bandwidth demonstrations have relied on benchtop equipment for filtering and modulating Kerr comb wavelength channels, data-centre interconnects require a compact on-chip form factor for these operations. Here we demonstrate a massively scalable chip-based silicon photonic data link using a Kerr comb source enabled by a new link architecture and experimentally show aggregate single-fibre data transmission of 512 Gb s −1 across 32 independent wavelength channels. The demonstrated architecture is fundamentally scalable to hundreds of wavelength channels, enabling massively parallel terabit-scale optical interconnects for future green hyperscale data centres.

Rizzo, Anthony (ORCID:000000034752797X)↗

Toucan: A performance portable, scalable implementation of the DECA algorithm

In the field of additive manufacturing (AM), cellular automata (CA) is extensively used to simulate microstructural evolution during solidification. However, while traditional CA approaches are relatively fast, they still require a substantial number of time steps, are limited to moderate volumes, and are relatively difficult to improve through parallelism due to the highly localized nature of the solidification front. Here, to address these issues of time to solution and load balancing, we introduce Toucan, a parallel, performance-portable, and scalable code written in C++ with the Kokkos library that leverages the discrete event inspired cellular automata (DECA) algorithm to perform parallel-in-time (PinT) grain growth simulations. Toucan effectively mitigates load balancing issues by distributing the computational workload more evenly across processors, enhancing scalability and efficiency. We conduct both strong and weak scaling studies on up to 64 GPUs on the Frontier supercomputer, demonstrating that Toucan significantly outperforms the current state-of-the-art, time-stepped CA code, ExaCA, on both single and multi-GPU simulations. Even in AM-specific weak scaling scenarios, Toucan maintains near-ideal scaling, in contrast to the linear increase observed with ExaCA due to the moving laser raster pattern. This study highlights Toucan’s potential to transform microstructural simulations in AM by radically improving both efficiency and scalability over existing methods.

36 MATERIALS SCIENCE↗

Virtual Flow Simulator

Virtual Flow Simulator (VFS) is a state-of-the-art computational fluid mechanics (CFD) package that is capable of simulating multi-physics/multi-phase flows with the most advanced turbulence models (RANS, LES) over complex terrains. The flow solver is based on the Curvilinear Immersed Boundary (CURVIB) method to handle geometrically complex and moving domains. Different modules of the VFS package can provide different simulation capabilities for specific applications ranging from the fluid-structure interaction (FSI) of solid and deformable bodies, the two-phase free surface flow solver based on the level set method for ocean waves, sediment transport models in rivers and the large-scale models of wind farms based on actuator lines and surfaces. All numerical features of VFS package have been validated with known analytical and experimental data as reported in the related journal articles. VFS package is suitable for a broad range of engineering applications within different industries. VFS has been used in different projects with applications in wind and hydrokinetic energy, offshore and near-shore ocean studies, cardiovascular and biological flows, and natural streams and river morphodynamics. Over the last decade, the development of VFS has been supported and assisted with the help of various United States companies and federal agencies that are listed in the sponsor lists. In this version, VFS-Wind contains all the necessary modeling tools for wind energy applications, including land-based and offshore wind farms. VFS is highly scalable to run on either desktop computers or high performance clusters (up to 16,000 CPUs). This released version comes with a detailed user’s manual and a set of case studies designed to facilitate the learning of the various aspects of the code in a comprehensive manner. The included documentation and support material has been elaborated in a collaboration effort with Sandia National Labs under the contract DE-EE0005482. The VFS-Wind source code is distributed under the GNU General Public License (GPL 2.0), offering to the community the flexibility to use and expand the capabilities of the package. We note that future publications including academic journals that result from the use of VFS must be cited appropriately under the GPL v.2 license agreement.

Yang, Xiaolei↗

Virtual Flow Simulator

Virtual Flow Simulator (VFS) is a state-of-the-art computational fluid mechanics (CFD) package that is capable of simulating multi-physics/multi-phase flows with the most advanced turbulence models (RANS, LES) over complex terrains. The flow solver is based on the Curvilinear Immersed Boundary (CURVIB) method to handle geometrically complex and moving domains. Different modules of the VFS package can provide different simulation capabilities for specific applications ranging from the fluid-structure interaction (FSI) of solid and deformable bodies, the two-phase free surface flow solver based on the level set method for ocean waves, sediment transport models in rivers and the large-scale models of wind farms based on actuator lines and surfaces. All numerical features of VFS package have been validated with known analytical and experimental data as reported in the related journal articles. VFS package is suitable for a broad range of engineering applications within different industries. VFS has been used in different projects with applications in wind and hydrokinetic energy, offshore and near-shore ocean studies, cardiovascular and biological flows, and natural streams and river morphodynamics. Over the last decade, the development of VFS has been supported and assisted with the help of various United States companies and federal agencies that are listed in the sponsor lists. In this version, VFS-Wind contains all the necessary modeling tools for wind energy applications, including land-based and offshore wind farms. VFS is highly scalable to run on either desktop computers or high performance clusters (up to 16,000 CPUs). This released version comes with a detailed user’s manual and a set of case studies designed to facilitate the learning of the various aspects of the code in a comprehensive manner. The included documentation and support material has been elaborated in a collaboration effort with Sandia National Labs under the contract DE-EE0005482. The VFS-Wind source code is distributed under the GNU General Public License (GPL 2.0), offering to the community the flexibility to use and expand the capabilities of the package. We note that future publications including academic journals that result from the use of VFS must be cited appropriately under the GPL v.2 license agreement.

Calderer, Antoni↗

A Study of Parallel Scalability and Dynamic Workload Balancing in GlennICE

The Glenn Icing Computational Environment (GlennICE) is a computational tool designed to calculate ice growth on complex three-dimensional geometries. It utilizes user-supplied computational fluid dynamics solutions for the geometry of interest. Key developments include advancements in convergence of collection efficiency, trajectory optimization, and refinement methodology. These improvements have significantly enhanced GlennICE’s efficiency for practical engineering applications. A recent study focused on benchmarking GlennICE’s scalability in a parallel environment using static scheduling. Findings indicated a potential twofold increase in efficiency through workload balance enhancements. This paper presents an analysis of the solver’s new workload balancing improvements, incorporating shared memory and dynamic scheduling routines. Results demonstrate a highly efficient and consistent algorithm across high-performance computing clusters.

Computational Icing↗

A Study of Parallel Scalability and Dynamic Workload Balancing in GlennICE

The Glenn Icing Computational Environment (GlennICE) is a computational tool designed to calculate ice growth on complex three-dimensional geometries. It utilizes user-supplied computational fluid dynamics solutions for the geometry of interest. Key developments include advancements in convergence of collection efficiency, trajectory optimization, and refinement methodology. These improvements have significantly enhanced GlennICE’s efficiency for practical engineering applications. A recent study focused on benchmarking GlennICE’s scalability in a parallel environment using static scheduling. Findings indicated a potential twofold increase in efficiency through workload balance enhancements. This paper presents an analysis of the solver’s new workload balancing improvements, incorporating shared memory and dynamic scheduling routines. Results demonstrate a highly efficient and consistent algorithm across high-performance computing clusters.

Computational Icing↗

Early Career Award: An Enabling Computational Framework for Uncertainty Assimilation and Propagation in Complex PDE Systems: Sparse and Low-rank Techniques (Final Report)

This project aims to develop a set of new theories, algorithms, and computational tools that enable ultra-scalable representation and propagation of high-dimensional uncertainties to circumvent the issue of curse-of-dimensionality. The research components are motivated by the fact that highly scalable algorithms are needed to achieve exascale performance for uncertainty quantification and predictive simulation at extreme scales. Although these developments are generic and applicable to a wide range of high-dimensional, complex, stochastic, partial-differential-equation (PDE) systems, this research effort will focus on predictive simulation of high-energy-density lithium batteries. To create knowledge about the precise failure mechanisms of these batteries and ultimately predict their performance and reliability, a truly data-driven, multi-scale, multi-physics framework is needed that allows studying (i) dominant sources of uncertainty in the governing physical models at all scales of relevance and (ii) the interface between the interchanging physics and scales. The impact of the planned mathematical and numerical advancements in prediction of high-dimensional uncertain systems will be broad across areas of engineering and science such as climate modeling, fusion energy, combustion, and nanoscience, among others.

97 MATHEMATICS AND COMPUTING↗

Atomistic non-adiabatic dynamics of the LH2 complex with a GPU-accelerated ab initio exciton model

We recently outlined an efficient multi-tiered parallel ab initio excitonic framework that utilizes time dependent density functional theory (TDDFT) to calculate ground and excited state energies and gradients of large supramolecular complexes in atomistic detail – enabling us to undertake non-adiabatic simulations which explicitly account for the coupled anharmonic vibrational motion of all the constituent atoms in a supramolecular system. Here we apply that framework to the 27 coupled bacterio-chlorophyll-a chromophores which make up the LH2 complex, using it to compute an on-the-fly nonadiabatic surface-hopping (SH) trajectory of electronically excited LH2. Part one of this article is focussed on calibrating our ab initio exciton Hamiltonian using two key parameters: a shift δ, which corrects for the error in TDDFT vertical excitation energies; and an effective dielectric constant ε, which describes the average screening of the transition-dipole coupling between chromophores. Using snapshots obtained from equilibrium molecular dynamics simulations (MD) of LH2, we tune the values of both δ and ε through fitting to the thermally broadened experimental absorption spectrum, giving a linear absorption spectrum that agrees reasonably well with experiment. In part two of this article, we construct a time-resolved picture of the coupled vibrational and excitation energy transfer (EET) dynamics in the sub-picosecond regime following photo-excitation. Assuming Franck–Condon excitation of a narrow eigenstate band centred at 800 nm, we use surface hopping to follow a single nonadiabatic dynamics trajectory within the full eigenstate manifold. Consistent with experimental data, this trajectory gives timescales for B800→B850 population transfer (τ B800→B850 ) between 650–1050 fs, and B800 population decay (τ 800→ ) between 10–50 fs. The dynamical picture that emerges is one of rapidly fluctuating LH2 eigenstates that are delocalized over multiple chromophores and undergo frequent crossing on a femtosecond timescale as a result of the atomic vibrations of the constituent chromophores. The eigenstate fluctuations arise from disorder that is driven by vibrational dynamics with multiple characteristic timescales. The scalability of our ab initio excitonic computational framework across massively parallel architectures opens up the possibility of addressing a wide range of questions, including how specific dynamical motions impact both the pathways and efficiency of electronic energy-transfer within large supramolecular systems.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

A Testbed of Parallel Kernels for Computer Science Research

For several decades, computer scientists have sought guidance on how to evolve architectures, languages, and programming models for optimal performance, efficiency, and productivity. Unfortunately, this guidance is most often taken from the existing software/hardware ecosystem. Architects attempt to provide micro-architectural solutions to improve performance on fixed binaries. Researchers tweak compilers to improve code generation for existing architectures and implementations, and they may invent new programming models for fixed processor and memory architectures and computational algorithms. In today's rapidly evolving world of on-chip parallelism, these isolated and iterative improvements to performance may miss superior solutions in the same way gradient descent optimization techniques may get stuck in local minima. In an initial study, we have developed an alternate approach that, rather than starting with an existing hardware/software solution laced with hidden assumptions, defines the computational problems of interest and invites architects, researchers and programmers to implement novel hardware/ software co-designed solutions. Our work builds on the previous ideas of computational dwarfs, motifs, and parallel patterns by selecting a representative set of essential problems for which we provide: An algorithmic description; scalable problem definition; illustrative reference implementations; verification schemes. For simplicity, we focus initially on the computational problems of interest to the scientific computing community but proclaim the methodology (and perhaps a subset of the problems) as applicable to other communities. We intend to broaden the coverage of this problem space through stronger community involvement. Previous work has established a broad categorization of numerical methods of interest to the scientific computing, in the spirit of the NAS Benchmarks, which pioneered the basic idea of a 'pencil and paper benchmark' in the 1990s. The initial result of the more modern study was the seven dwarfs, which was subsequently extended to 13 motifs. These motifs have already been useful in defining classes of applications for architecture-software studies. However, these broad-brush problem statements often miss the nuance seen in individual kernels. For example, the computational requirements of particle methods vary greatly between the naive (but more accurate) direct calculations and the particle-mesh and particle-tree codes. Thus we commenced our study with an enumeration of problems, but then proceeded by providing not only reference implementations for each problem, but more importantly a mathematical definition that allows one to escape iterative approaches to software/hardware optimization. To ensure long term value, we have augmented each of our reference implementations with both a scalable problem generator and a verification scheme. In a paper we have prepared that documents our efforts, we describe in detail this process of problem definition, scalable input creation, verification, and implementation of reference codes for the scientific computing domain. Table 1 enumerates and describes the level of support we've developed for each kernel. We group these important kernels using the Berkeley dwarfs/motifs taxonomy using a red box in the appropriate column. As kernels become progressively complex, they build upon other, simpler computational methods. We note this dependency via orange boxes. After enumeration of the important numerical problems, we created a domain-appropriate high-level definition of each problem. To ensure future endeavors are not tainted by existing implementations, we specified the problem definition to be independent of both computer architecture and existing programming languages, models, and data types. Then, to provide context as to how such kernels productively map to existing architectures, languages and programming models, we produced reference implementations for most of the kernel (see table). These sample codes should be viewed as 'hints,' designed to show how other designers have mapped a problem's operands and operators to existing hardware and software. Since we wanted such implementations to be illustrative, we tried to ensure they were the most straightforward implementation in the easiest to understand languages using familiar architectures. To that end, most of the linear algebra-oriented computations are written in MATLAB using array indexing to process matrices, rather than one-line library calls to compute the same kernel. This ensures that the kernel's computation is explicit and readable in the implementation and not hidden behind a library. For other problems, such as the Barnes-Hut n-Body solver, the implementations were written in pure C, without any supporting library computations. Along this line, we have also created a scalable problem generator to accompany each computation.

97 MATHEMATICS AND COMPUTING↗

Petascale Data Storage Institute (Final Report)

pNFS is an extension to NFSv4 that helps clients overcome NFS scalability and performance barriers. Like NFS, pNFS is a client/server protocol implemented with secure and reliable remote procedure calls. pNFS departs from conventional NFS by allowing clients to access storage directly and in parallel. This helps overcome server bottlenecks inherent to NAS access methods. Making pNFS available to petascale researchers required coordinated progress in several dimensions, a confluence of multiple processes that took nearly ten years. The protocol had to be specified and published by the IETF. Nascent implementations had to track changes in the draft specification, changes in the Linux kernel, and interoperate with one another. The process by which modifications are accepted into the Linux kernel by maintainers and developers (and ultimately by Linus Torvalds himself) itself required consensus and compromise. Finally, pNFS support needed to be provided by the major Linux distributors (Red Hat, SUSE, etc.) and storage vendors (Netapp, EMC, IBM, Microsoft, etc.) on both the engineering and product sides. The specification requirement was met in December 2008, when the IETF approved the NFSv4.1 specification, which incorporates pNFS and pNFS file layouts, as a Proposed Standard as well as the (separate) specifications of pNFS object and block layouts. Throughout the PDSI project, CITI was the key contributor to the Linux-based, open source implementation of NFSv4.1 and pNFS. Considerable effort was devoted to the process of refining the IETF specification, which underwent dozens of preliminary drafts, rebasing implementations to the latest Linux kernel, which itself changed quarterly, and to testing interoperability with other developers. At last, the Linux pNFS implementation was incorporated into the Linux mainline kernel, although this was achieved after the PDSI project ended.

97 MATHEMATICS AND COMPUTING↗

A scalable approach to solving dense linear algebra problems on hybrid CPU-GPU systems

Aiming to fully exploit the computing power of all CPUs and all graphics processing units (GPUs) on hybrid CPU-GPU systems to solve dense linear algebra problems, in this paper we design a class of heterogeneous tile algorithms to maximize the degree of parallelism, to minimize the communication volume, and to accommodate the heterogeneity between CPUs and GPUs. The new heterogeneous tile algorithms are executed upon our decentralized dynamic scheduling runtime system, which schedules a task graph dynamically and transfers data between compute nodes automatically. The runtime system uses a new distributed task assignment protocol to solve data dependencies between tasks without any coordination between processing units. By overlapping computation and communication through dynamic scheduling, we are able to attain scalable performance for the double-precision Cholesky factorization and QR factorization. Finally, our approach demonstrates a performance comparable to Intel MKL on shared-memory multicore systems and better performance than both vendor (e.g., Intel MKL) and open source libraries (e.g., StarPU) in the following three environments: heterogeneous clusters with GPUs, conventional clusters without GPUs, and shared-memory systems with multiple GPUs.

97 MATHEMATICS AND COMPUTING↗

Reducing Communication in Algebraic Multigrid Using Additive Variants

Algebraic multigrid (AMG) has proven to be an effective scalable solver on many high performance computers. However, its increasing communication complexity on coarser levels has shown to seriously impact its performance on computers with high communication cost. Moreover, additive AMG variants provide not only increased parallelism as well as decreased numbers of messages per cycle but also generally exhibit slower convergence. Here we present various new additive variants with convergence rates that are significantly improved compared to the classical additive algebraic multigrid method and investigate their potential for decreased communication, and improved communication-computation overlap, features that are essential for good performance on future exascale architectures.

97 MATHEMATICS AND COMPUTING↗

GPU Acceleration of the Locally Selfconsistent Multiple Scattering Code for First Principles Calculation of the Ground State and Statistical Physics of Materials

The Locally Self-consistent Multiple Scattering (LSMS) code solves the first principles Density Functional theory Kohn-Sham equation for a wide range of materials with a special focus on metals, alloys and metallic nano-structures. It has traditionally exhibited near perfect scalability on massively parallel high performance computer architectures. We present our efforts to exploit GPUs to accelerate the LSMS code to enable first principles calculations of O(100,000) atoms and statistical physics sampling of finite temperature properties. Using the Cray XK7 system Titan at the Oak Ridge Leadership Computing Facility we achieve a sustained performance of 14.5PFlop/s and a speedup of 8.6 compared to the CPU only code.

Eisenbach, Markus↗

Stable parallel training of Wasserstein conditional generative adversarial neural networks

In this work, we propose a stable, parallel approach to train Wasserstein conditional generative adversarial neural networks (W-CGANs) under the constraint of a fixed computational budget. Differently from previous distributed GANs training techniques, our approach avoids inter-process communications, reduces the risk of mode collapse and enhances scalability by using multiple generators, each one of them concurrently trained on a single data label. The use of the Wasserstein metric also reduces the risk of cycling by stabilizing the training of each generator. We illustrate the approach on the CIFAR10, CIFAR100, and ImageNet1k datasets, three standard benchmark image datasets, maintaining the original resolution of the images for each dataset. Performance is assessed in terms of scalability and final accuracy within a limited fixed computational time and computational resources. To measure accuracy, we use the inception score, the Fréchet inception distance, and image quality. An improvement in inception score and Fréchet inception distance is shown in comparison to previous results obtained by performing the parallel approach on deep convolutional conditional generative adversarial neural networks as well as an improvement of image quality of the new images created by the GANs approach. Weak scaling is attained on both datasets using up to 2000 NVIDIA V100 GPUs on the OLCF supercomputer Summit.

97 MATHEMATICS AND COMPUTING↗

Inverse design of hypoeutectoid pearlite steel microstructures using a deep learning and genetic algorithm optimization framework

Goal-oriented microstructure design in metallic materials is a challenging task due to complex structure-property relationships. Traditional experimental and computational approaches are time-intensive and economically inefficient, limiting their applicability for large-scale design space exploration. Here, in this work, we propose an end-to-end framework that integrates deep learning models with genetic optimization to design microstructures with targeted mechanical properties. Deep learning models enable accurate forward design, while their integration with genetic optimization enables efficient inverse design within a few hours, compared to days or weeks using conventional finite element simulations. The framework combines experimental characterization and finite element modeling to analyze the influence of microstructural features on the mechanical behavior of hypoeutectoid steels. Data from both experiments and simulations are used to train the deep learning models. To demonstrate its effectiveness, we apply the framework to 0.63% carbon steel with proeutectoid ferrite and pearlite phases, commonly used in industrial applications. In this study, 2D microstructures were used for modeling, selected primarily for computational efficiency and to establish proof of concept. The framework successfully optimizes microstructures for targeted yield strength, ultimate strength, and stress concentration factors while significantly reducing computational time. Beyond hypoeutectoid steels, this scalable framework can be extended to other material systems and integrated with additive manufacturing, offering an efficient approach for accelerating microstructure design for specific engineering applications.

ConvLSTM↗