Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “workload”

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 19 records
All sourcesNASA NTRSDOE OSTI
DOE OSTI2023

$\mathrm{RADICAL}$-Pilot and $\mathrm{PMIx}$/$\mathrm{PRRTE}$: Executing Heterogeneous Workloads at Large Scale on Partitioned $\mathrm{HPC}$ Resources

Execution of heterogeneous workflows on high-performance computing (HPC) platforms present unprecedented resource management and execution coordination challenges for runtime systems. Task heterogeneity increases the complexity of resource and execution management, limiting the scalability and efficiency of workflow execution. Re-source partitioning and distribution of tasks execution over portioned re-sources promises to address those problems but we lack an experimental evaluation of its performance at scale. Here this paper provides a performance evaluation of the Process Management Interface for Exascale (PMIx) and its reference implementation PRRTE on the leadership-class HPC plat-form Summit, when integrated into a pilot-based runtime system called RADICAL-Pilot. We partition resources across multiple PRRTE Distributed Virtual Machine (DVM) environments, responsible for launching tasks via the PMIx interface. We experimentally measure the work-load execution performance in terms of task scheduling/launching rate and distribution of DVM task placement times, DVM startup and termination overheads on the Summit leadership-class HPC platform. Integrated solution with PMIx/PRRTE enables using an abstracted, standardized set of interfaces for orchestrating the launch process, dynamic process management and monitoring capabilities. It extends scaling capabilities allowing to overcome a limitation of other launching mechanisms (e.g., JSM/LSF). Explored different DVM setup configurations provide insights on DVM performance and a layout to leverage it. Our experimental results show that heterogeneous workload of 65,500 tasks on 2048 nodes, and partitioned across 32 DVMs, runs steady with resource utilization not lower than 52%. While having less concurrently executed tasks resource utilization is able to reach up to 85%, based on results of heterogeneous workload of 8200 tasks on 256 nodes and 2 DVMs.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Comparative evaluation of deep learning workloads for leadership-class systems

Deep learning (DL) workloads and their performance at scale are becoming important factors to consider as we design, develop and deploy next-generation high-performance computing systems. Since DL applications rely heavily on DL frameworks and underlying compute (CPU/GPU) stacks, it is essential to gain a holistic understanding from compute kernels, models, and frameworks of popular DL stacks, and to assess their impact on science-driven, mission-critical applications. At Oak Ridge Leadership Computing Facility (OLCF), we employ a set of micro and macro DL benchmarks established through the Collaboration of Oak Ridge, Argonne, and Livermore (CORAL) to evaluate the AI readiness of our next-generation supercomputers. In this paper, we present our early observations and performance benchmark comparisons between the Nvidia V100 based Summit system with its CUDA stack and an AMD MI100 based testbed system with its ROCm stack. We take a layered perspective on DL benchmarking and point to opportunities for future optimizations in the technologies that we consider.

Yin, Junqi↗
DOE OSTI2024

Integrating the PanDA Workload Management System with the Vera C. Rubin Observatory

The Vera C. Rubin Observatory will produce an unprecedented astronomical data set for studies of the deep and dynamic universe. Its Legacy Survey of Space and Time (LSST) will image the entire southern sky every three to four days and produce tens of petabytes of raw image data and associated calibration data over the course of the experiment’s run. More than 20 terabytes of data must be stored every night, and annual campaigns to reprocess the entire dataset since the beginning of the survey will be conducted over ten years. The Production and Distributed Analysis (PanDA) system was evaluated by the Rubin Observatory Data Management team and selected to serve the Observatory’s needs due to its demonstrated scalability and flexibility over the years, for its Directed Acyclic Graph (DAG) support, its support for multi-site processing, and its highly scalable complex workflows via the intelligent Data Delivery Service (iDDS). PanDA is also being evaluated for prompt processing where data must be processed within 60 seconds after image capture. This paper will briefly describe the Rubin Data Management system and its Data Facilities (DFs). Finally, it will describe in depth the work performed in order to integrate the PanDA system with the Rubin Observatory to be able to run the Rubin Science Pipelines using PanDA.

79 ASTRONOMY AND ASTROPHYSICS↗
DOE OSTI2023

Neck Muscle Coactivation Response to Varied Levels of Mental Workload During Simulated Flight Tasks

Objective To evaluate neck muscle coactivation across different levels of mental workload during simulated flight tasks. Background Neck pain (NP) is highly prevalent among military aviators. Given the complex nature within the flight environment, mental workload may be a risk factor for NP. This may induce higher levels of neck muscle coactivity, which over time may accelerate fatigue, increase neck discomfort, and affect flight task performance. Method Three counterbalanced mental workload conditions represented by simulated flight tasks modulated by interstimulus frequency and complexity were investigated using the Modifiable Multitasking Environment (ModME). The primary measure was a neck coactivation index to describe the neuromuscular effort of the neck muscles as a system. Additional measures included perceived workload (NASA TLX), subjective discomfort, and task performance. Participants ( n = 60; 30M, 30F) performed three test conditions over 1 hr each while seated in a simulated seating environment. Results Neck coactivation indices (CoA) and subjective neck discomfort corresponded with increasing level of mental workload. Average CoAs for low, medium, and high workloads were: .0278(SD = .0232), .0286(SD = .0231), and .0295(SD = .0228), respectively. NASA TLX mental, temporal, effort, and overall scores also increased with the level of mental workload assigned. For ModME task performance, the overall performance score, monitoring accuracy, and resource management accuracy decreased while reaction times increased with the increasing level of mental workload. Communication accuracy was lowest with the low mental workload but had higher reaction times relative to increasing workload. Conclusion Mental workload affects neck muscle coactivation during combinations of simulated flight tasks within a simulated helicopter seating environment. Application The results of this study provide insights into the physical response to mental workload. With increasing multisensory modalities within the work environment, these insights may assist the consideration of physical effects from cognitive factors.

Behavioral Sciences↗
DOE OSTI2023

Micromovements and discomfort associated with flight mission with helmet operation tasks with different levels of cognitive workload

When performing stationary tasks under elevated cognitive workload, individuals must perform continual muscle contractions to maintain stability of the body, resulting in fatigue of the postural muscles. When the muscles perform these contractions in a prolonged manner, the body potentially responds through small changes in body movements—micromovements that may lead to discomfort. The study purpose was to evaluate impact of cognitive load on micromovements. The micromovements were measured during three different cognitive workloads; low, medium, and high. The NASA-TLX score was used to evaluate the perceived mental workload and discomfort was assessed by visual analog scale. In total, 60 subjects (30 males and 30 females) were recruited and performed cognitive tasks that simulated flight operations such as changing the radio frequency based on air traffic control messages, balancing the fuel levels in simulated fuel tanks, and aiming a reticle in a designated moving target using the cyclic control. Cognitive load was defined by the frequency of events. Micromovements were defined by changes in the center of pressure (COP) of the seat pan and COP standard deviation. It was found that the high cognitive workloads had the highest NASA-TLX scores including mental demands, temporal demands, and effort. The neck area had the highest overall levels of discomfort followed by upper back. The highest standard deviation for COP shift and number of micromovements occurred for medium cognitive workloads. In conclusion, while there were some interesting trends, few trends reached a statistical significance due to high variability among subjects for the outcome variables.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2020

Union: An Automatic Workload Manager for Accelerating Network Simulation

With the rapid growth of the machine learning applications, the workloads of future HPC systems are anticipated to be a mix of scientific simulation, big data analytics, and machine learning applications. Simulation is a great research vehicle to understand the performance implications of co-running scientific applications with big data and machine learning workloads on large-scale systems. In this paper, we present Union, a workload manager that provides an automatic framework to facilitate hybrid workload simulation in CODES. Furthermore, we use Union, along with CODES, to investigate various hybrid workloads composed of traditional simulation applications and emerging learning applications on two dragonfly systems. The experiment results show that both message latency and communication time are important performance metrics to evaluate network interference. Network interference on HPC applications is more reflected by the message latency variation, whereas ML application performance depends more on the communication time.

heterogeneous workloads↗
DOE OSTI2026

Dataset of Generative AI Workload Power Profiles

This dataset provides a collection of high-resolution (5/10 Hz or every 0.2/0.1 seconds) power consumption profiles for generative artificial intelligence (GenAI) workloads executed on NLR's High Performance Computing (HPC) platform Kestrel. The dataset also includes examples of representative whole-facility power profiles generated using a bottom-up, event-driven, data center energy model . This dataset is designed to support research in energy modeling, infrastructure planning, energy system integration, and sustainability analysis for AI-driven computing systems. The dataset captures time-resolved electrical power measurements across a diverse set of configurations, including variations in job type (inference vs. training), workload (LLM vs. image generation), datasets, and number of compute nodes. Power traces are provided in a standardized format and include both raw/instantaneous and aggregated files. Each profile is accompanied by metadata describing workload parameters, enabling reproducibility and cross-study comparison. The dataset is intended for use in applications such as data center infrastructure planning, energy modeling, demand response and grid impact studies, and development and validation of system-level simulation tools. By making these workload-specific power profiles publicly available, this dataset aims to address the current lack of open, empirical energy data for generative AI systems and to facilitate transparent, reproducible research on the energy and environmental impacts of large-scale AI deployment. If you use this dataset, please cite the associated publication: Vercellino et al., “Measurement of Generative AI Workload Power Profiles for Whole-Facility Data Center Infrastructure Planning,” arXiv:2604.07345 (2026).

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2021

Characterizing Machine Learning I/O Workloads on Leadership Scale HPC Systems

High performance computing (HPC) is no longer solely limited to traditional workloads such as simulation and modeling. With the increase in the popularity of machine learning (ML) and deep learning (DL) technologies, we are observing that an increasing number of HPC users are incorporating ML methods into their workflow and scientific discovery processes, across a wide spectrum of science domains such as biology, earth science, and physics. This gives rise to a diverse set of I/O patterns than the traditional checkpoint/restart-based HPC I/O behavior. The details of the I/O characteristics of such ML I/O workloads have not been studied extensively for large-scale leadership HPC systems. This paper aims to fill that gap by providing an in-depth analysis to gain an understanding of the I/O behavior of ML I/O workloads using darshan - an I/O characterization tool designed for lightweight tracing and profiling. We study the darshan logs of more than 23, 000 HPC ML I/O jobs over a time period of one year running on Summit - the second-fastest supercomputer in the world. This paper provides a systematic I/O characterization of ML I/O jobs running on a leadership scale supercomputer to understand how the I/O behavior differs across science domains and the scale of workloads, and analyze the usage of parallel file system and burst buffer by ML I/O workloads.

Paul, Arnab↗
DOE OSTI2019

Optimization of Workload Distribution of Data Centers Based on a Self-Learning In Situ Adaptive Tabulation Model

Data center cooling typically involves non-uniform airflow and temperature distributions, which are affected by the IT workload distribution. It is helpful to simulate the airflow and temperature to optimize the workload distribution. Traditional computational fluid dynamics (CFD) simulation is usually time-consuming while conventional reduced order models (ROMs), though computationally fast, may generate inaccurate results even after being fully trained. In Situ Adaptive Tabulation (ISAT), contracting to conventional ROM, can make prediction with error lower than a user-specified tolerance. To demonstrate using of ISAT for optimal workload distribution in data center, this paper presents a preliminary study of an ISAT-based genetic algorithm optimization platform. The ISAT is trained offline by using the results from CFD simulations using a hypothetical simple data center. The optimal workload distribution determined by the platform leads to approximately 6.8% of energy savings when compared to the benchmark with a uniform workload distribution. We note that the time cost for the entire optimization process, including the training of ISAT is about 4 hours, which is acceptable in the design phase.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗
DOE OSTI2025

The water use of data center workloads: A review and assessment of key determinants

The global importance of data center water use is increasing with the rapid growth of digitalization and artificial intelligence. This study analyzes the factors influencing workload-level water use, measured in liters consumed per workload, to guide water-saving strategies in data centers. Our findings reveal workload-level water use variations exceeding 10,000-fold, driven by over 1000-fold differences in water consumption per kilowatt hour of server electricity consumed and approximately 10-fold differences in server workload efficiency. Key determinants are ranked as server efficiency, electrical grid water consumption factors, server utilization, cooling system type, infrastructure efficiency, climate zone, inactive server percentage, and server refresh cycle. Notably, there is no single recipe for minimizing water use; instead, optimal outcomes depend on tailored combinations of these factors. This analysis addresses critical knowledge gaps by identifying the determinants of data center water use and exploring their achievable minima under diverse site-specific constraints.

Data centers↗
DOE OSTI2020

Indicator-directed Dynamic Power Management for Iterative Workloads on GPU-Accelerated Systems

Modern high-performance and warehouse computing centers show strong interest in minimizing system power consumption while satisfying customers’ quality of service (QoS). Dynamic voltage and frequency scaling (DVFS) is effective for achieving this goal. Nevertheless, automating the process online and making it transparent to users must address three major challenges: (1) Complexity — today’s hardware components (e.g., CPUs, GPUs, memory, network, etc.) can be configured in several or dozens of frequency/voltage states for satisfying divergent system demands. Given their combination and the emergence of heterogeneity, searching the optimal configuration in the design space online can be timing consuming. (2) QoS guarantee — user-defined objectives such as power constraint and performance target must be monitored, predicted and ensured at the best effort. (3) Adaptability — various known and unknown workloads run on systems. Workloads characteristics should be quickly determined and configurations dynamically adjusted in accord with workloads and QoS. In this work, we focus on applications exhibiting an interesting feature – iterative or periodic, which is common among conventional HPC and emerging machine learning workloads. We propose an online dynamic power-performance (ODPP) management framework to dynamically adjust GPU DVFS configurations to meet performance and power objectives and constraints, without any code annotation or intrusion. Particularly, ODPP extracts the performance and power indicators for applications from their resources utilization profiles in a short episode. It further automatically constructs an accurate model that infers from the indicators how the application's performance and power vary with GPU core and memory frequencies. Aided with the model, for both seen and unseen applications, ODPP can quickly determine the most appropriate DVFS configuration for their execution. We evaluate ODPP on an NVIDIA GPU using multiple exascale computing (ECP) and deep learning applications.

Zou, Pengfei↗
DOE OSTI2021

Parallel I/O Evaluation Techniques and Emerging HPC Workloads: A Perspective

Emerging workloads such as artificial intelligence, big data analytics and complex multi-step workflows alongside future exascale applications are anticipated future HPC workloads, which will result in a more diverse I/O system workload and even less predictable I/O behavior and access patterns. Along with the ever increasing gap between the compute and storage performance capabilities, the in-depth understanding of extreme-scale I/O behavior and the I/O performance modeling and prediction are essential tools of the large-scale I/O evaluation process for addressing the needs of extreme-scale hybrid workloads. In this survey article, we focus on the state-of-the-art of the I/O behavior and performance analysis process for HPC systems in a 5-year time window and identify future research challenges.

Neuwirth, Sarah↗
DOE OSTI2019

Exploring the Psychometrics of Common Post-Scenario Human Factors Questionnaires of Workload, Situation Awareness, and Perceived Difficulty

Abstract: The United States (U.S.) Department of Energy (DOE) Light Water Reactor Sustainability (LWRS) program has the primary mission of sustaining operation of the existing commercial nuclear power plants (NPPs) through a multi-pathway approach in conducting research and development (R&D). In this effort, the role of human factors engineering (HFE) has been an integral in ensuring new digital upgrades that are part of the overall main control room modernization strategy for various U.S. NPPs optimize human-system performance and do not introduce new human failure modes. In this process, testing and evaluation of concept (i.e., prototype) human machine interfaces (HMIs) are done in operator-in-the-loop studies at various stages of the HFE modification process, such as during design and verification and validation (V&V) [1]. During these studies, licensed operators go through select scenarios using a testbed, such as a full-scale simulator or at a NPP’s training simulator, to identify potential usability issues and provide general design feedback regarding the new digital HMIs. Human-system performance is measured and evaluated through a range of objective and self-report measures that capture aspects of plant performance, task performance, workload, situation awareness, and anthropometric and ergonomic concerns [2]. The use of post-scenario questionnaires is an acceptable method for gathering measures of workload and situation awareness. To this end, the NASA-Task Load Index (TLX) [e.g., 3, 4, & 5] and Situation Awareness Rating Technique (SART) [e.g., 6 & 7] are common standardized questionnaires used for evaluating workload and situation awareness, respectively. When combined, these questionnaires consist of a total of 15 questions, each using a different rating scale. Debriefs are also done through interviews with operators to gather design insight after each scenario. Since operator- and training simulator-time is limited, a concern during this post-scenario period of operator-in-the-loop studies is the limited time available to run scenarios, administer post-scenario questionnaires, and debrief within the allotted simulator time. To address this potential problem, this paper explores the psychometric characteristics of NASA-TLX, SART, and a common one-question post-scenario usability questionnaire (i.e., the single ease question – SEQ [8]) in effort to inform future development of a condensed questionnaire that captures aspects of workload and situation awareness. This exploratory work uses data from four different operator-in-the-loop studies completed under prior LWRS operator-in-the-loop studies. Findings from this paper should support future development of a consolidated questionnaire that may be used in operator-in-the-loop studies, when time is compressed, that conforms to the HFE scientific and regulatory requirements set for NPP control room modernization.

99 GENERAL AND MISCELLANEOUS↗
DOE OSTI2021

Checkpoint/Restart Vision and Strategies for NERSC’s Production Workloads

As a primary approach to fault-tolerant computing, Checkpoint/Restart (C/R) improves scientific productivity for users, provides scheduling flexibility for computing centers, and protects against system failures. While both applicationspecific (or application-level) and transparent C/R are used in practice, we are interested in transparent checkpointing, which is vital for system-level checkpointing. Developing and maintaining transparent C/R tools for HPC applications, however, is labor intensive and highly complex due to ever-changing HPC systems and diverse production workloads. Existing C/R tools are often research-oriented, so there is a gap to close before they can be used reliably with production workloads, especially on cutting edge HPC systems. In this position paper, we present our journey to prepare a production-ready MPI-Agnostic Network-Agnostic (MANA) transparent checkpointing tool for NERSC, and share our vision and strategies to bring transparent C/R capabilities to NERSC’s production workloads on current and future systems.

42 ENGINEERING↗
DOE OSTI2019

Domain-specific virtual processors as a portable programming and execution model for parallel computational workloads on modern heterogeneous high-performance computing architectures

We advocate domain-specific virtual processors (DSVP) as a portability layer for expressing and executing domain-specific computational workloads on modern heterogeneous HPC architectures, with applications in quantum chemistry. Specifically, in this article we extend, generalize and better formalize the concept of a domain-specific virtual processor as applied to scientific high-performance computing. In particular, we introduce a system-wide recursive (hierarchical) hardware encapsulation mechanism into the DSVP architecture and specify a concrete microarchitectural design of an abstract DSVP from which specialized DSVP implementations can be derived for specific scientific domains. Furthermore, we demonstrate, an example of a domain-specific virtual processor specialized to numerical tensor algebra workloads, which is implemented in the ExaTENSOR library developed by the author with a primary focus on the quantum many-body computational workloads on large-scale GPU-accelerated HPC platforms.

97 MATHEMATICS AND COMPUTING↗