Search NASA⌕ Search

SEARCH · Search NASA

Results for “Heterogeneous computing architectures”

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 55 records · Page 3
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Advanced Computing Annual Report 2023

In 2023, advanced computing saw the arrival of Kestrel, the National Renewable Energy Laboratory's (NREL's) newest high-performance computing (HPC) system. Kestrel will accelerate clean energy research at a pace and scale more than five times greater than Eagle, with approximately 44 petaflops of computing power. Kestrel's heterogeneous architecture - which includes both CPU-only and GPU-accelerated nodes - is designed to bring a much greater GPU capacity to EERE workloads compared to Eagle, enabling rapidly advancing applications in artificial intelligence and expanding research in new directions for computing. In Fiscal Year (FY) 2023, 333 projects utilized NREL's HPC system, advancing the U.S. Department of Energy's (DOE's) Office of Energy Efficiency and Renewable Energy (EERE) mission across 13 funding areas. Cross-disciplinary collaboration among researchers yielded more than 800 technical outputs, including 177 peer-reviewed journal articles in FY 2023. All this great work continues to advance the science of energy efficiency and renewable energy. This report highlights research that utilized HPC resources in FY 2023.

advanced computing↗
DOE OSTI2022

High Performance Adaptive Physics Refinement to Enable Large-Scale Tracking of Cancer Cell Trajectory

The ability to track simulated cancer cells through the circulatory system, important for developing a mechanistic understanding of metastatic spread, pushes the limits of today's supercomputers by requiring the simulation of large fluid volumes at cellular-scale resolution. To overcome this challenge, we introduce a new adaptive physics refinement (APR) method that captures cellular-scale interaction across large domains and leverages a hybrid CPU-GPU approach to maximize performance. Through algorithmic advances that integrate multi-physics and multi-resolution models, we establish a finely resolved window with explicitly modeled cells coupled to a coarsely resolved bulk fluid domain. In this work we present multiple validations of the APR framework by comparing against fully resolved fluid-structure interaction methods and employ techniques, such as latency hiding and maximizing memory bandwidth, to effectively utilize heterogeneous node architectures. Collectively, these computational developments and performance optimizations provide a robust and scalable framework to enable system-level simulations of cancer cell transport.

59 BASIC BIOLOGICAL SCIENCES↗
DOE OSTI2023

Validation of time-dependent shift using the pulsed sphere benchmarks

The detailed behavior of neutrons in a rapidly changing time-dependent physical system is a challenging computational physics problem, particularly when using Monte Carlo methods on heterogeneous high-performance computing architectures. A small number of algorithms and code implementations have been shown to be performant for time-independent (fixed source and k-eigenvalue) Monte Carlo, and there are existing simulation tools that successfully solve the time-dependent Monte Carlo problem on smaller computing platforms. To bridge this gap, a time-dependent version of ORNL’s Shift code has been recently developed. Shift’s history-based algorithm on CPUs, and its event-based algorithm on GPUs, have both been observed to scale well to very large numbers of processors, which motivated the extension of this code to solve time-dependent problems. The validation of this new capability requires a comparison with time-dependent neutron experiments. Lawrence Livermore National Laboratory’s (LLNL) pulsed sphere benchmark experiments were simulated in Shift to validate both the time-independent as well as new time-dependent features recently incorporated into Shift. A suite of pulsed-sphere models was simulated using Shift and compared to the available experimental data and simulations with MCNP. Overall results indicate that Shift accurately simulates the pulsed sphere benchmarks, and that the new time-dependent modifications of Shift are working as intended. Validated exascale neutron transport codes are essential for a wide variety of future multiphysics applications.

Palmer, Camille J.↗
NASA NTRSDate not supplied

Enabling Execution of a Legacy CFD Mini Application on Accelerators Using OpenMP

We describe the process and outcome of our efforts to port a legacy Fortran benchmark code to heterogeneous GPU-accelerated computing architectures using OpenMP. The benchmark code is one of the multi-zone NAS Parallel Benchmarks (NPB-MZ) called SP-MZ. This “mini-app” mimics the computation and data movement that is found in popular legacy and modern implicit computational fluid dynamics (CFD)solvers. Our objective was to examine how efficiently legacy Fortran codes can be ported to accelerators by leveraging OpenMP directives. We describe the development and optimization process and demonstrate the performance impact of various code modifications. We show select profiling results from the Nvidia nvvp profiler to help others diagnose and overcome performance issues in their own applications. We present results for two compute systems endowed with Nvidia V100 accelerators.

Ioannis Nompelis↗
DOE OSTI2023

A PIC Bootstrapping Strategy for Exascale CFD-DEM Simulation

The exascale computing era comes with the release of MFIX-Exa, a new code for studying gas-particle fluidization using CFD-DEM and PIC models on massively parallel, heterogeneous high-performance computing architectures. However, as compute resources continue to grow in scale and complexity, so too does the impetus to use them efficiently. Here, we propose a novel bootstrapping method to minimize neglected simulation time and cast the approach more broadly among a growing field of multi-fidelity methods.

Porcu, Roberto↗
DOE OSTI2026

The ocean model for E3SM global applications: Omega version 0.1.0 – a new high-performance computing code for exascale architectures

This paper introduces Omega, the Ocean Model for E3SM Global Applications. Omega is a new ocean model designed to run efficiently on high performance computing (HPC) platforms, including exascale heterogeneous architectures with accelerators, such as Graphics Processing Units (GPUs). Omega is written in C and uses the Kokkos performance portability library. These were chosen because they are well-supported and will help future-proof Omega for upcoming HPC architectures. Omega will eventually replace the Model for Prediction Across Scales-Ocean (MPAS-Ocean) in the US Department of Energy's (DOE's) Energy Exascale Earth System Model (E3SM). Omega runs on unstructured horizontal meshes with variable-resolution capability and implements the same horizontal discretization as MPAS-Ocean. This work documents the design and performance of Omega Version 0.1.0 (Omega-V0), which solves the shallow water equations with passive tracers and is the first step towards the full primitive equation ocean model. On Central Processing Units (CPUs), Omega-V0 is 1.4 times faster than MPAS-Ocean with the same configuration. Omega-V0 is more efficient on GPUs than CPUs on a per-watt basis – by a factor of 5.3 on Frontier and 3.6 on Aurora, two of the world's fastest exascale computers.

54 ENVIRONMENTAL SCIENCES↗
NASA NTRS1995

DREAMS and IMAGE: A Model and Computer Implementation for Concurrent, Life-Cycle Design of Complex Systems

Computing architectures are being assembled that extend concurrent engineering practices by providing more efficient execution and collaboration on distributed, heterogeneous computing networks. Built on the successes of initial architectures, requirements for a next-generation design computing infrastructure can be developed. These requirements concentrate on those needed by a designer in decision-making processes from product conception to recycling and can be categorized in two areas: design process and design information management. A designer both designs and executes design processes throughout design time to achieve better product and process capabilities while expanding fewer resources. In order to accomplish this, information, or more appropriately design knowledge, needs to be adequately managed during product and process decomposition as well as recomposition. A foundation has been laid that captures these requirements in a design architecture called DREAMS (Developing Robust Engineering Analysis Models and Specifications). In addition, a computing infrastructure, called IMAGE (Intelligent Multidisciplinary Aircraft Generation Environment), is being developed that satisfies design requirements defined in DREAMS and incorporates enabling computational technologies.

Hale, Mark A.↗
NASA NTRS1994

An incremental database access method for autonomous interoperable databases

We investigated a number of design and performance issues of interoperable database management systems (DBMS's). The major results of our investigation were obtained in the areas of client-server database architectures for heterogeneous DBMS's, incremental computation models, buffer management techniques, and query optimization. We finished a prototype of an advanced client-server workstation-based DBMS which allows access to multiple heterogeneous commercial DBMS's. Experiments and simulations were then run to compare its performance with the standard client-server architectures. The focus of this research was on adaptive optimization methods of heterogeneous database systems. Adaptive buffer management accounts for the random and object-oriented access methods for which no known characterization of the access patterns exists. Adaptive query optimization means that value distributions and selectives, which play the most significant role in query plan evaluation, are continuously refined to reflect the actual values as opposed to static ones that are computed off-line. Query feedback is a concept that was first introduced to the literature by our group. We employed query feedback for both adaptive buffer management and for computing value distributions and selectivities. For adaptive buffer management, we use the page faults of prior executions to achieve more 'informed' management decisions. For the estimation of the distributions of the selectivities, we use curve-fitting techniques, such as least squares and splines, for regressing on these values.

Roussopoulos, Nicholas↗
NASA NTRS1993

Node assignment in heterogeneous computing

A number of node assignment schemes, both static and dynamic, are explored for the Algorithm to Architecture Mapping Model (ATAMM). The architecture under consideration consists of heterogeneous processors and implements dataflow models of real-time applications. Terminology is developed for heterogeneous computing. New definitions are added to the ATAMM for token and assignment classifications. It is proved that a periodic execution is possible for dataflow graphs. Assignment algorithms are developed and proved. A design procedure is described for satisfying an objective function in an heterogeneous architecture. Several examples are provided for illustration.

Som, Sukhamoy↗
DOE OSTI2026

HARMONY: Large-Scale Architecture Search for Efficient Hybrid Language Models

As large language models scale to trillions of parameters, their computational and memory requirements present critical challenges for efficient training and deployment. While Mixture of Experts (MoE) architectures enable efficient scaling through sparse parameter activation, and state-space models like Mamba offer linear-time complexity, principled methods for combining these paradigms remain undeveloped. We introduce HARMONY (Hybrid Architecture Research for Mamba, Optimized with Neural efficiencY), a multi-objective evolutionary neural architecture search framework for discovering efficient hybrid language models that integrate Transformer attention mechanisms, Mixture-of-Experts routing, and Mamba state-space components. Through large-scale distributed search using 16,384 MI250X GPUs on the Frontier supercomputer, HARMONY explores a comprehensive design space encompassing six attention variants (MHA, MQA, GQA, MLA, SWA, and Mamba-2), variable MoE configurations with both routed and shared experts, and extensive Mamba hyperparameters. Our framework discovers heterogeneous architectures that balance training performance with computational efficiency through multi-objective optimization incorporating latency penalties and fitness-based selection. Analysis of discovered architectures reveals that optimal hybrid designs favor heterogeneous component mixing rather than homogeneous patterns, with Mamba-2 and Multi-Head Latent Attention (MLA) emerging as preferred mechanisms. Discovered architectures demonstrate superior training efficiency: our best configuration achieves a final perplexity of 1.0874 with 2.38B parameters while processing 4,320 tokens/second, outperforming significantly larger manually designed models. Full-scale evaluation shows HARMONY's top architectures achieve better loss trajectories than equivalently-sized models using state-of-the-art configurations including Mixtral, Jamba, and Samba. Additionally, we demonstrate 91% weak scaling efficiency when training discovered 36B-parameter models across 1,024 GPUs. HARMONY is released as an open framework with comprehensive tools for building and training hybrid models using expert-data-pipeline parallelism, democratizing access to automated architecture design for next-generation language models.

Herron, Emily [ORNL] (ORCID:0000000273008172)↗
NASA NTRS1995

Using an architectural approach to integrate heterogeneous, distributed software components

Many computer programs cannot be easily integrated because their components are distributed and heterogeneous, i.e., they are implemented in diverse programming languages, use different data representation formats, or their runtime environments are incompatible. In many cases, programs are integrated by modifying their components or interposing mechanisms that handle communication and conversion tasks. For example, remote procedure call (RPC) helps integrate heterogeneous, distributed programs. When configuring such programs, however, mechanisms like RPC must be used explicitly by software developers in order to integrate collections of diverse components. Each collection may require a unique integration solution. This paper describes improvements to the concepts of software packaging and some of our experiences in constructing complex software systems from a wide variety of components in different execution environments. Software packaging is a process that automatically determines how to integrate a diverse collection of computer programs based on the types of components involved and the capabilities of available translators and adapters in an environment. Software packaging provides a context that relates such mechanisms to software integration processes and reduces the cost of configuring applications whose components are distributed or implemented in different programming languages. Our software packaging tool subsumes traditional integration tools like UNIX make by providing a rule-based approach to software integration that is independent of execution environments.

Callahan, John R.↗
DOE OSTI2020

Pre-exascale accelerated application development: The ORNL Summit experience

High-performance computing (HPC) increasingly relies on heterogeneous architectures to achieve higher performance. In the Oak Ridge Leadership Facility (OLCF), Oak Ridge, TN, USA, this trend continues as its latest supercomputer, Summit, entered production in early 2019. The combination of IBM POWER9 CPU and NVIDIA V100 GPU, along with a fast NVLink2 interconnect and other latest technologies, pushes system performance to a new height and breaks the exascale barrier by certain measures. Due to Summit's powerful GPUs and much higher GPU–CPU ratio, offloading to accelerators becomes a requirement for any application, which intends to effectively use the system. To facilitate navigating a complex landscape of competing heterogeneous architectures, a collection of applications from a wide spectrum of scientific domains is selected for early adoption on Summit. In this article, the experience and lessons learned are summarized, in the hope of providing useful guidance to address new programming challenges, such as scalability, performance portability, and software maintainability, for future application development efforts on heterogeneous HPC systems.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Optimizing Error-Bounded Lossy Compression for Scientific Data on GPUs

Error-bounded lossy compression is a critical technique for significantly reducing scientific data volumes. With ever-emerging heterogeneous high-performance computing (HPC) architecture, GPU-accelerated error-bounded compressors (such as CUSZ and cuZFP) have been developed. However, they suffer from either low performance or low compression ratios. To this end, we propose CUSZ+ to target both high compression ratios and throughputs. We identify that data sparsity and data smoothness are key factors for high compression throughputs. Our key contributions in this work are fourfold: (1) We propose an efficient compression workflow to adaptively perform run-length encoding and/or variable-length encoding. (2) We derive Lorenzo reconstruction in decompression as multidimensional partial-sum computation and propose a fine-grained Lorenzo reconstruction algorithm for GPU architectures. (3) We carefully optimize each of CUSZ kernels by leveraging state-of-the-art CUDA parallel primitives. (4) We evaluate CUSZ+ using seven real-world HPC application datasets on V100 and A100 GPUs. Experiments show CUSZ+ improves the compression throughputs and ratios by up to 18.4x and 5.3x, respectively, over CUSZ on the tested datasets.

Tian, Jiannan↗
NASA NTRS1994

Accommodating Heterogeneity in a Debugger for Distributed Computations

In an ongoing project at NASA Ames Research Center, we are building debugger for distributed computations running on a heterogeneous set of machines. Historically, such debuggers have been built as front-ends to existing source-level debuggers on the target platforms. In effect, these back-end debuggers are providing a collection of debugger services to a client. The major drawback is that because of inconsistencies among the back-end debuggers, the front-end must use a different protocol when talking to each back-end debugger. This can make the front-end quite complex. We have avoided this complexity problem by defining the client-server debugger protocol. While it does require vendors to adapt their existing debugger code to meet the protocol, vendors are generally interested in doing so because the approach has several advantages. In addition to solving the heterogenous platform debugging problem, it will be pos~sible to write interesting debugger user interfaces that can be easily ported across a variety of machines. This will likely encourage investment in application-domain specific debuggers. In fact, the user interface of our debugger will be geared to scientists developing computational fluid dynamics codes. This paper describes some of the problems encountered in developing a portable debugger for heterogenous, distributed computing and how the architecture of our debugger avoids them. It then provides a detailed description of the debugger client-server protocol. Some of the more interesting attributes of the protocol are: (1) It is object-oriented; (2) It uses callback functions to capture the asynchronous nature of debugging in a procedural fashion; (3) It contains abstractions, such as in-line instrumentation, for the debugging of computationally intensive programs; (4) For remote debugging, it has operations that enable the implementor to optimize message passing traffic between client and server. The soundness of the protocol is being tested through the implementation of a prototype client and server pair.

Hood, Robert↗
DOE OSTI2021

Coding the Computing Continuum: Fluid Function Execution in Heterogeneous Computing Environments

Advances in network technologies have greatly decreased barriers to accessing physically distributed computers. This newfound accessibility coincides with increasing hardware specialization, creating exciting new opportunities to dispatch workloads to the best resource for a specific purpose, rather than those that are closest or most easily accessible. We present Delta, a service designed to intelligently schedule function-based workloads across a distributed set of heterogeneous computing resources. Delta implements an extensible architecture in which different predictors and scheduling algorithms can be integrated to provide dynamically evolving estimates of function execution times on different resources-estimates that can be used to determine the most appropriate location for execution. We describe predictors for function runtime, data transfer time, and cold-start resource provisioning and configuration delay; dynamic learning methods that update predictor models over time; and scheduling strategies that take into account both function and endpoint information. We show that these methods can halve workload makespan when compared with a strategy that selects the fastest resource, and decrease makespan by a factor of five when compared to a round robin strategy, when deployed on a heterogeneous testbed with resources ranging from a Raspberry Pi to a GPU node in an academic cloud.

Computing continuum↗
DOE OSTI2025

SYCL for Performance Portability: Application Experience with Coupled Cluster Formalism in Quantum Chemistry on Exascale Systems

The exascale computing has brought unprecedented heterogeneity in node architectures, with systems such as Frontier and Aurora featuring diverse GPU accelerators, network connectivity among others. Ensuring performance portability across these platforms is a key challenge. To address this, we employ the SYCL programming model to develop portable, high-performance quantum chemistry workloads. As a representative application, we focus on the non-iterative Triples component of the coupled-cluster CCSD(T) method, a key driver in quantum chemistry. In this work, we report on our experience deploying SYCL-based implementations using both DPC++ and AdaptiveCPP across two flagship exascale platforms: OLCF Frontier with AMD MI250X GPUs and ALCF Aurora with Intel GPUs. Our results demonstrate that SYCL enables efficient, single-source implementations that scale to thousands of nodes, delivering performance on par with vendor-optimized HIP solutions. We highlight key insights into runtime behavior, kernel portability, and scaling characteristics, showing that SYCL offers a viable path for performance-portable computing.

Bagusetty, Abhishek [Argonne National Laboratory (↗
DOE OSTI2023

Large-Scale Materials Modeling at Quantum Accuracy: Ab Initio Simulations of Quasicrystals and Interacting Extended Defects in Metallic Alloys

Ab initio electronic-structure has remained dichotomous between achievable accuracy and length-scale. Quantum many-body (QMB) methods realize quantum accuracy but fail to scale. Density functional theory (DFT) scales favorably but remains far from quantum accuracy. We present a framework that breaks this dichotomy by use of three interconnected modules: (i) invDFT: a methodological advance in inverse DFT linking QMB methods to DFT; (ii) MLXC: a machine-learned density functional trained with invDFT data, commensurate with quantum accuracy; (iii) DFT-FE-MLXC: an adaptive higher-order spectral finite-element (FE) based DFT implementation that integrates MLXC with efficient solver strategies and HPC innovations in FE-specific dense linear algebra, mixed-precision algorithms, and asynchronous compute-communication. Furthermore, we demonstrate a paradigm shift in DFT that not only provides an accuracy commensurate with QMB methods in ground-state energies, but also attains an unprecedented performance of 659.7 PFLOPS (43.1% peak FP64 performance) on 619,124 electrons using 8,000 GPU nodes of Frontier supercomputer.

density functional theory↗
DOE OSTI2019

Sierra Center of Excellence: Lessons Learned

The introduction of heterogeneous computing via GPUs from the Sierra architecture represented a significant shift in direction for computational science at Lawrence Livermore National Laboratory (LLNL), and therefore required significant preparation. Over the last five years, the Sierra Center of Excellence (CoE) has brought employees with specific expertise from IBM and NVIDIA together with LLNL in a concentrated effort to prepare applications, system software, and tools for the Sierra supercomputer. This article shares the process we applied for the CoE and documents lessons learned during the collaboration, with the hope that others will be able to learn from both our success and intermediate setbacks. We describe what we have found to work for the management of such a collaboration and best practices for algorithms and source code, system configuration and software stack, tools, and application performance.

97 MATHEMATICS AND COMPUTING↗