Search NASA⌕ Search

SEARCH · Search NASA

Results for “GPU accelerators”

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 883 records · Page 49

Scalable One‐Step Gel Conversion Route to High‐Performance CHA Zeolite Hollow Fiber Membranes and Modules for CO 2 Separation

Efficient separation of CO 2 from other molecules such as CH 4 , N 2 , and hydrocarbons is a key problem in many energy production and greenhouse gas reduction schemes. Zeolite materials demonstrate attractive gas separation properties, but their industrial application as membranes in gas separation are elusive because of high cost and low reproducibility of conventional fabrication techniques. A facile method is demonstrated for the fabrication of high‐silica CHA zeolite membranes on low‐cost alumina hollow fiber substrates by a one‐step direct gel conversion method, along with a detailed characterization of the CHA hollow fiber membranes. To demonstrate the scalability of the method, 20 hollow fiber (710 μm OD and 390 μm ID) membranes are synthesized in the same batch and assembled into a module. Under dry conditions, the membranes show an excellent CO 2 permeance up to 3500 GPU (1 Gas Permeation Unit = 3.348 × 10 −10 mol m −2 s −1 Pa −1 ) and an excellent CO 2 separation performance (e.g., CO 2 /CH 4 separation factors up to 210 in the 1–7 bar feed pressure range). In addition, the membranes exhibit a considerable increase in CO 2 selectivity under humid (saturated water vapor) conditions with moderate permeance drops. This study opens a promising new path for accelerated development of low‐cost zeolitic membranes for gas separation technology.

Yang, Shaowei↗

The uintah framework: a unified heterogeneous task scheduling and runtime system

The development of a new unified, multi-threaded runtime system for the execution of asynchronous tasks on heterogeneous systems is described in this work. These asynchronous tasks arise from the Uintah framework, which was developed to provide an environment for solving a broad class of fluid-structure interaction problems on structured adaptive grids. Uintah has a clear separation between its MPI-free user-coded tasks and its runtime system that ensures these tasks execute efficiently. This separation also allows for complete isolation of the application developer from the complexities involved with the parallelism Uintah provides. While we have designed scalable runtime systems for large CPU core counts, the emergence of heterogeneous systems, with additional on-node accelerators and co-processors presents additional design challenges in terms of effectively utilizing all computational resources on-node and managing multiple levels of parallelism. Our work addresses these challenges for Uintah by the development of new hybrid runtime system and Unified multi-threaded MPI task scheduler, enabling Uintah to fully exploit current and emerging architectures with support for asynchronous, out-of-order scheduling of both CPU and GPU computational tasks. This design coupled with an approach that uses MPI to communicate between nodes, a shared memory model on-node and the use of novel lock-free data structures, has made it possible for Uintah to achieve excellent scalability for challenging fluid-structure problems using adaptive mesh refinement on as many as 256K cores on the DoE Jaguar XK6 system. This design has also demonstrated an ability to run capability jobs on the heterogeneous systems, Keeneland and TitanDev. In this work, the evolution of Uintah and its runtime system is examined in the context of our new Unified multi-threaded scheduler design. The performance of the Unified scheduler is also tested against previous Uintah scheduler and runtime designs over a range of processor core and GPU counts.

Source record↗

Algorithmic Patterns for $$\mathcal {H}$$-Matrices on Many-Core Processors

In this work, we consider the reformulation of hierarchical ( \(\mathcal {H}\) ) matrix algorithms for many-core processors with a model implementation on graphics processing units (GPUs). \(\mathcal {H}\) matrices approximate specific dense matrices, e.g., from discretized integral equations or kernel ridge regression, leading to log-linear time complexity in dense matrix–vector products. The parallelization of \(\mathcal {H}\) matrix operations on many-core processors is difficult due to the complex nature of the underlying algorithms. While previous algorithmic advances for many-core hardware focused on accelerating existing \(\mathcal {H}\) matrix CPU implementations by many-core processors, we here aim at totally relying on that processor type. As main contribution, we introduce the necessary parallel algorithmic patterns allowing to map the full \(\mathcal {H}\) matrix construction and the fast matrix–vector product to many-core hardware. In this work, crucial ingredients are space filling curves, parallel tree traversal and batching of linear algebra operations. The resulting model GPU implementation hmglib is the, to the best of the authors knowledge, first entirely GPU-based Open Source \(\mathcal {H}\) matrix library of this kind. We investigate application examples as present in kernel ridge regression, Gaussian Process Regression and kernel-based interpolation. In this context, an in-depth performance analysis and a comparative performance study against a standard multi-core CPU \(\mathcal {H}\) matrix library highlights profound speedups of our many-core parallel approach.

97 MATHEMATICS AND COMPUTING↗

Viskores: Integrating Parallel Scientific Visualization Research into Applications

Viskores is a scientific visualization library that is the primary deployment of such algorithms to the parallel accelerated processors of modern DOE supercomputers. In this paper, we review the capabilities provided by Viskores and how these capabilities are leveraged by other software in the high-performance computing ecosystem. We discuss the Viskores data representation and pay particular attention to array management. Through this array management we describe how data is adapted between Viskores and other software along with strategies for converting dynamic, polymorphic objects to static representations better suited to GPU processing. We conclude with several examples of Viskores integrating with high-performance software that is used in production today.

Moreland, Ken [ORNL] (ORCID:0000000270513288)↗

Applying Machine‐Learning Methods to Laser Acceleration of Protons: Lessons Learned From Synthetic Data

ABSTRACT In this study, we consider three different machine‐learning methods—a three‐hidden‐layer neural network, support vector regression, and Gaussian process regression—and compare how well they can learn from a synthetic data set for proton acceleration in the Target Normal Sheath Acceleration regime. The synthetic data set was generated from a previously published theoretical model by Fuchs et al. 2005 that we modified. Once trained, these machine‐learning methods can assist with efforts to maximize the peak proton energy, or with the more general problem of configuring the laser system to produce a proton energy spectrum with desired characteristics. In our study, we focus on both the accuracy of the machine‐learning methods and the performance on one GPU including memory consumption. Although it is arguably the least sophisticated machine‐learning model we considered, support vector regression performed very well in our tests.

Desai, Ronak↗

BSTC: A Novel Binarized-Soft-Tensor-Core Design for Accelerating Bit-Based Approximated Neural Nets

Binarized neural network (or BNN) promises tremendous performance improvement over traditional DNNs through simplified bit-level computation and significantly reduced memory access/storage cost. In addition, it has other advantages of low-cost, low-energy and high-robustness, showing great utilization potential in resources-constrained, volatile and latency-critical applications, which are critical for future HPC execution. However, the promised significant performance gain of BNN inference has never been fully demonstrated on general-purpose processors, particularly on GPUs, due to: (i) the challenge to extract and leverage sufficient fine-grained bit-level-parallelism to saturate GPU cores when batch is small; (ii) the fundamental design conflict between bit-based BNN algorithm and underlying word-based architecture; (iii) architecture \& performance unfriendly BNN network design. To address (i) and (ii), we propose binarized-soft-tensor-core as a software-hardware codesign approach to construct bit-manipulation capability for modern GPUs to effectively harvest the emerging bit-level-parallelism. To tackle (iii), we propose intra- and inter-layer fusion techniques so that the entire BNN inference can be packed into a single GPU kernel, to avoid high-cost frequent launching. Experiments demonstrate that our design can achieve over 1000x speedup for raw inference latency and 10x for inference throughput over the state-of-the-art full-precision simulated BNN inference for AlexNet on ImageNet.

Li, Ang↗

Bridging Cloud and Edge Computing at NREL Using CONNECT: Cloud Optimized Networking for Next-Gen Edge Computing Technologies [Slides]

CONNECT is an innovative on-premise hardware and software solution that integrates edge and cloud computing infrastructure at NREL. Built on the AWS Greengrass middleware and leveraging the MQTT protocol, CONNECT enables real-time data streaming from IoT devices and gateways to both cloud and local services, empowering researchers to rapidly capture, analyze, and act upon edge-generated data while leveraging cloud capabilities. The platform addresses research infrastructure challenges by providing a pre-approved platform which is already configured with the correct networking and cybersecurity baselines thus eliminating procurement delays and enabling on-demand availability. CONNECT's hybrid architecture efficiently manages burstable workloads, allowing research teams to dynamically scale computational capacity, handle peak data loads, and reduce operational bottlenecks. Advanced capabilities include built-in GPU support for executing machine learning models which enables low-latency inference at the edge from models trained in the cloud. This architecture supports real-time analytics and filtering, providing a mechanism to allow only transmitting and processing high-value data. Cloud-based configuration management permits engineers to manage on-premise systems remotely, optimizing operational efficiency. By bridging edge and cloud computing, CONNECT provides NREL researchers with a flexible, scalable platform that accelerates scientific discovery while maintaining robust security and performance standards.

97 MATHEMATICS AND COMPUTING↗

Multicore and Accelerator Development for a Leadership-Class Stellar Astrophysics Code

We describe recent development work on the core-collapse supernova code CHIMERA. CHIMERA has consumed more than 100 million cpu-hours on Oak Ridge Leadership Computing Facility (OLCF) platforms in the past 3 years, ranking it among the most important applications at the OLCF. Most of the work described has been focused on exploiting the multicore nature of the current platform (Jaguar) via, e.g., multithreading using OpenMP. In addition, we have begun a major effort to marshal the computational power of GPUs with CHIMERA. The impending upgrade of Jaguar to Titan a 20+ PF machine with an NVIDIA GPU on many nodes makes this work essential.

Messer, Bronson↗

Acceleration of Graph Neural Network-Based Prediction Models in Chemistry via Co-Design Optimization on Intelligence Processing Units

Atomic structure prediction and associated property calculations are the bedrock of chemical physics. Since high-fidelity ab initio modeling techniques for computing the structure and properties can be prohibitively expensive, this motivates the development of machine-learning (ML) models that make these predictions more efficiently. Training graph neural networks over large atomistic databases introduces unique computational challenges such as the need to process millions of small graphs with variable size and support communication patterns that are distinct from learning over large graphs such as social networks. We demonstrate a novel hardware-software co-design approach to scale up the training of atomistic graph neural networks (GNN) for structure and property prediction. First, to eliminate redundant computation and memory associated with alternative padding techniques and to improve throughput via minimizing communication, we formulate the effective coalescing of the batches of variable-size atomistic graphs as the bin packing problem and introduce a hardware-agnostic algorithm to pack these batches. In addition, we propose hardware-specific optimizations including a planner and vectorization for the gather-scatter operations targeted for Graphcore’s Intelligence Processing Unit (IPU), as well as model-specific optimizations such as merged communication collectives and optimized softplus. Putting these all together, we demonstrate the effectiveness of the proposed co-design approach by providing an implementation of a well-established atomistic GNN on the Graphcore IPUs. We evaluate the training performance on multiple atomistic graph databases with varying degrees of graph counts, sizes and sparsity. Here, we demonstrate that such a co-design approach can reduce the training time of atomistic GNNs and can improve the performance by up to 1.5× compared to the baseline implementation of the model on the IPUs. Additionally, we compare our IPU implementation with a Nvidia GPU-based implementation and show that our atomistic GNN implementation on the IPUs can run 1.8× faster on average compared to the execution time on the GPUs.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

SU-E-T-29: A Web Application for GPU-Based Monte Carlo IMRT/VMAT QA with Delivered Dose Verification

Purpose: To enable an existing web application for GPU-based Monte Carlo (MC) 3D dosimetry quality assurance (QA) to compute “delivered dose” from linac logfile data. Methods: We added significant features to an IMRT/VMAT QA web application which is based on existing technologies (HTML5, Python, and Django). This tool interfaces with python, c-code libraries, and command line-based GPU applications to perform a MC-based IMRT/VMAT QA. The web app automates many complicated aspects of interfacing clinical DICOM and logfile data with cutting-edge GPU software to run a MC dose calculation. The resultant web app is powerful, easy to use, and is able to re-compute both plan dose (from DICOM data) and delivered dose (from logfile data). Both dynalog and trajectorylog file formats are supported. Users upload zipped DICOM RP, CT, and RD data and set the expected statistic uncertainty for the MC dose calculation. A 3D gamma index map, 3D dose distribution, gamma histogram, dosimetric statistics, and DVH curves are displayed to the user. Additional the user may upload the delivery logfile data from the linac to compute a 'delivered dose' calculation and corresponding gamma tests. A comprehensive PDF QA report summarizing the results can also be downloaded. Results: We successfully improved a web app for a GPU-based QA tool that consists of logfile parcing, fluence map generation, CT image processing, GPU based MC dose calculation, gamma index calculation, and DVH calculation. The result is an IMRT and VMAT QA tool that conducts an independent dose calculation for a given treatment plan and delivery log file. The system takes both DICOM data and logfile data to compute plan dose and delivered dose respectively. Conclusion: We sucessfully improved a GPU-based MC QA tool to allow for logfile dose calculation. The high efficiency and accessibility will greatly facilitate IMRT and VMAT QA.

07 ISOTOPE AND RADIATION SOURCES↗

Performance Portable Graphics Processing Unit Acceleration of a High-Order Finite Element Multiphysics Application

The Lawrence Livermore National Laboratory (LLNL) will soon have in place the El Capitan exascale supercomputer, based on advanced micro devices (AMD) graphics processing units (GPUs). As part of a multiyear effort under the National Nuclear Security Administration (NNSA) Advanced Simulation and Computing (ASC) program, we have been developing marbl, a next generation, performance portable multiphysics application based on high-order finite elements. In previous years, we successfully ported the Arbitrary Lagrangian–Eulerian (ALE), multimaterial, compressible flow capabilities of marbl to nvidia GPUs as described in Vargas et al. Here, in this paper, we describe our ongoing effort in extending marbl's GPU capabilities with additional physics, including multigroup radiation diffusion and thermonuclear burn for high energy density physics (HEDP) and fusion modeling. We also describe how our portability abstraction approach based on the raja Portability Suite and the mfem finite element discretization library has enabled us to achieve high performance on AMD based GPUs with minimal effort in hardware-specific porting. Throughout this work, we highlight numerical and algorithmic developments that were required to achieve GPU performance.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Evaluating the Performance of Integer Sum Reduction in SYCL on GPUs

SYCL is a promising programming model for heterogeneous computing—allowing a single-source code to target devices from multiple vendors. One significant task performed on these accelerators is a primitive operation for integer sum reduction. This paper presents several SYCL implementations of integer sum reduction—using atomic functions, shared local memory, vectorized memory accesses and parameterized workload sizes—to compare the performance and maturity of SYCL against open-source vendor-specific implementations of the same reduction. For a sufficiently large number of integers, tuning the parameters of our SYCL implementations achieves 1.4X speedup over the open-source implementations on an Intel UHD630 integrated GPU. The SYCL reduction is 3% faster than the templated reduction in Thrust, and 0.3% faster than the device reduction in CUB on an Nvidia P100 GPU. The SYCL reduction is 1.9% faster than the templated reduction in Thrust, and 0.4% faster than the device reduction in CUB on an Nvidia V100 GPU.

Jin, Zheming↗

kynema-fmb [SWR-23-07]

Kynema-FMB (FKA: Kynema) is an open-source performance portable flexible multibody (FMB) dynamics solver designed for time-domain simulations. While originally tailored for wind turbine structural dynamics, the formulation and implementation are those of a general flexible-multidbody dynamics solver that can readily be applied to a wide range of systems. Kynema was designed with a narrow focus, namely to provide a lightweight, fast, accurate FMD solver for coupling to computational-fluid-dynamics (CFD) codes, especially the CFD codes in the Kynema suite, for fluid-structure-interaction (FSI) simulations. Kynema-FMB is equipped to model systems that can be represented as a collection of beams and rigid bodies that are connected through constraints. Degrees of freedom are defined in the inertial/global frame of reference and include displacements and rotations (formally as rotation matrices, but stored as quaternions). The underlying formulation is built on a Lie-group time integrator designed for index-3 differential-algebraic equations, which is second-order accurate in time (Bruls et al., 2012). Beam models are based on geometrically exact beam theory and are discretized as high-order spectral finite elements similar to those in BeamDyn (Wang et al., 2017). The governing equations for a FMD system like a wind turbine constitute a highly nonlinear system of constrained partial-differential equations. Kynema-FMB uses analytical Jacobians in the nonlinear-system solves in each time step. Linear systems use sparse storage and several third-party sparse-linear-system solvers are enabled. Ill conditioning of linear systems is mitigated with preconditioning described in Bottasso et al, 2008. Kynema-FMB is integrated with a simple open-source controller (ROSCO). There is an application programming interface (API) for coupling to geometry-resolved CFD (like that in Sharma et al., 2023) and actuator-force CFD (like that in Kuhn et al., 2025). In the latter, for actuator-line models, Kynema-FMB includes an internal blade-element solver that depends on user-provided lookup tables for coefficients of lift and drag, i.e., aerodynamic polars. Kynema-FMB is written in C++ and leverages Kokkos and Kokkos-Kernels (KokkosEcosystem) as its performance portability layer enabling simulations on both CPU and GPU systems. The repository is equipped with extensive automated testing at the unit and regression/system levels. The following describes the high-level development objectives conceived for Kynema: *Kynema will follow modern software development best practices, including test-driven development (TDD), version control, hierarchical automated testing, and continuous integration (CI) for a robust development environment. *The core data structures are memory efficient and enable vectorization and parallelization at multiple levels. *Data structures are data-oriented to exploit methods for accelerated computing including high utilization of chip resources (e.g., single instruction multiple data (SIMD) instruction sets) and parallelization using GP-GPUs. *The computational algorithms incorporate robust open-source libraries for mathematical operations, resource allocation, and data management. *The API design considers multiple stakeholder needs and ensure integration with existing and future ecosystems for data science, machine learning, and AI. *Kynema-FMB is written in modern C++ and leverages Kokkos as its performance-portability library with inspiration from the kynema stack.

Sprague, MichaelA.↗

Accelerated Depth Computation for Surface Boxplots with Deep Learning

Functional depth is a well-known technique used to derive descriptive statistics (e.g., median, quartiles, and outliers) for 1D data. Surface boxplots extend this concept to ensembles of images, helping scientists and users identify representative and outlier images. However, the computational time for surface boxplots increases cubically with the number of ensemble members, making it impractical for integration into visualization tools. In this paper, we propose a deep-learning solution for efficient depth prediction and computation of surface boxplots for time-varying ensemble data. Our deep learning framework accurately predicts member depths in a surface boxplot, achieving average speedups of 6X on a CPU and 15X on a GPU for the 2D Red Sea dataset with 50 ensemble members compared to the traditional depth computation algorithm. Our approach achieves at least a 99% level of rank preservation, with order flipping occurring only at pairs with extremely similar depth values that pose no statistical differences. This local flipping does not significantly impact the overall depth order of the ensemble members.

Han, Mengjiao↗

Towards Batched Linear Solvers on Accelerated Hardware Platforms

As hardware evolves, an increasingly effective approach to develop energy efficient, high-performance solvers, is to design them to work on many small and independent problems. Indeed, many applications already need this functionality, especially for GPUs, which are known to be currently about four to five times more energy efficient than multicore CPUs for every floating-point operation. In this paper, we describe the development of the main one-sided factorizations: LU, QR, and Cholesky; that are needed for a set of small dense matrices to work in parallel. We refer to such algorithms as batched factorizations. Our approach is based on representing the algorithms as a sequence of batched BLAS routines for GPU-contained execution. Note that this is similar in functionality to the LAPACK and the hybrid MAGMA algorithms for large-matrix factorizations. But it is different from a straightforward approach, whereby each of GPU's symmetric multiprocessors factorizes a single problem at a time. We illustrate how our performance analysis together with the profiling and tracing tools guided the development of batched factorizations to achieve up to 2-fold speedup and 3-fold better energy efficiency compared to our highly optimized batched CPU implementations based on the MKL library on a two-sockets, Intel Sandy Bridge server. Compared to a batched LU factorization featured in the NVIDIA's CUBLAS library for GPUs, we achieves up to 2.5-fold speedup on the K40 GPU.

Haidar, Azzam↗

Many-Core Acceleration of a Discrete Ordinates Transport Mini-App at Extreme Scale, In: ISC High Performance 2016: High Performance Computing

Time-dependent deterministic discrete ordinates transport codes are an important class of application which provide significant challenges for large, many-core systems. One such challenge is the large memory capacity needed by the solve step, which requires us to have a scalable solution in order to have enough node-level memory to store all the data. In our previous work, we demonstrated the first implementation which showed a significant performance benefit for single node solves using GPUs. In this paper we extend our work to large problems and demonstrate the scalability of our solution on two Petascale GPU-based supercomputers: Titan at Oak Ridge and Piz Daint at CSCS. Our results show that our improved node-level parallelism scheme scales just as well across large systems as previous approaches when using the tried and tested KBA domain decomposition technique. We validate our results against an improved performance model which predicts the runtime of the main ‘sweep’ routine when running on different hardware, including CPUs or GPUs.

Computer Science↗

MatRIS: Addressing the Challenges for Portability and Heterogeneity Using Tasking for Matrix Decomposition (Cholesky)

The ubiquitous in-node heterogeneity of HPC and cloud computing platforms makes software portability and performance optimization extremely challenging. Described here, the MatRIS multilevel math library abstraction framework employs tasking to alleviate these difficulties. MatRIS includes the IRIS task-based runtime on the bottom level and exposes different layers of abstraction to render algorithms architecturally agnostic. MatRIS ensures the decomposition and creation of tasks that represent the necessary encapsulation of the optimized kernels from both vendor and open-source math libraries. Once built, MatRIS can select different combinations of accelerators at runtime, making it portable even on diverse heterogeneous architectures. By leveraging the IRIS runtime’s features for managing heterogeneity, MatRIS deploys algorithms that remove the need to specify orchestration and data transfer. This study describes how the serial task abstraction of a tiled Cholesky factorization is made portable and scalable in the case of multi-device and multi-vendor heterogeneity on a node with NVIDIA and AMD GPUs by using MatRIS. First, we demonstrate that Cholesky in MatRIS provides multi-GPU scalability that offers competitive performance versus cuSolverMG. Then, we present the challenges and opportunities for heterogeneous execution.

Monil, M. A. H.↗

CoreTSAR: Core Task-Size Adapting Runtime

Heterogeneity continues to increase at all levels of computing, with the rise of accelerators such as GPUs, FPGAs, and other co-processors into everything from desktops to supercomputers. As a consequence, efficiently managing such disparate resources has become increasingly complex. CoreTSAR seeks to reduce this complexity by adaptively worksharing parallel-loop regions across compute resources without requiring any transformation of the code within the loop. Lastly, our results show performance improvements of up to three-fold over a current state-of-the-art heterogeneous task scheduler as well as linear performance scaling from a single GPU to four GPUs for many codes. In addition, CoreTSAR demonstrates a robust ability to adapt to both a variety of workloads and underlying system configurations.

97 MATHEMATICS AND COMPUTING↗