SEARCH · Search NASA

Results for “Workload Analysis”

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.

SourceResearch connectionFrom yearThrough year

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 109 records · Page 6

All sources NASA NTRS DOE OSTI

DOE OSTI2026

ChatMPI: LLM-Driven MPI Code Generation for HPC Workloads

The Message Passing Interface (MPI) standard plays a crucial role in enabling scientific applications for parallel computing and is an essential component in high-performance computing (HPC). However, implementing MPI code manually—especially applying a proper domain decomposition and communication pattern—is a challenging and error-prone task. We present ChatMPI, an AI assistant for MPI parallelization of sequential C codes. In our analysis, we focus on testing six essential HPC workloads, which are based on Basic Linear Algebra Subprograms levels 1, 2, and 3 as well as sparse, stencil, and iterative operations. We analyze the process of creating ChatMPI by using the ChatHPC library. This lightweight large language model (LLM)–based infrastructure enables HPC experts to efficiently create and supervise trustworthy AI capabilities for critical HPC software tasks. We study the data required for training (fine-tuning) ChatMPI to generate parallel codes that not only use MPI syntax correctly but also apply HPC techniques to reduce memory communication and maximize performance by using proper work decomposition. With a relatively small training dataset composed of a few dozen prompts and fewer than 15 minutes of fine-tuning on one node equipped with two NVIDIA H100 GPUs, ChatMPI elevates trustworthiness for MPI code generation of current LLMs (e.g., Code Llama, ChatGPT-4o and ChatGPT 5). Additionally, we evaluate the performance of the MPI codes generated by ChatMPI in comparison with the ones generated by ChatGPT-4o and ChatGPT-5. The codes generated by ChatMPI provide up to a 4 × boost in performance by using better problem decomposition, communication patterns, and HPC techniques (e.g., communication avoiding).

Valero Lara, Pedro [ORNL] (ORCID:0000000214794310)↗

NASA NTRS2012

SARDA HITL Preliminary Human Factors Measures and Analyses

Human factors data collected during the SARDA HITL Simulation Experiment include a variety of subjective measures, including the NASA TLX, questionnaire questions regarding situational awareness, advisory usefulness, UI usability, and controller trust. Preliminary analysis of the TLX data indicate that workload may not be adversely affected by use of the advisories, additionally, the controller's subjective ratings of the advisories may suggest acceptance of the tool.

Hyashi, Miwa↗

NASA NTRS2001

Development and Validation of the Controller Acceptance Rating Scale (CARS): Results of Empirical Research

The measurement of operational acceptability is important for the development, implementation, and evolution of air traffic management decision support tools. The Controller Acceptance Rating Scale was developed at NASA Ames Research Center for the development and evaluation of the Passive Final Approach Spacing Tool. CARS was modeled after a well-known pilot evaluation rating instrument, the Cooper-Harper Scale, and has since been used in the evaluation of the User Request Evaluation Tool, developed by MITRE's Center for Advanced Aviation System Development. In this paper, we provide a discussion of the development of CARS and an analysis of the empirical data collected with CARS to examine construct validity. Results of intraclass correlations indicated statistically significant reliability for the CARS. From the subjective workload data that were collected in conjunction with the CARS, it appears that the expected set of workload attributes was correlated with the CARS. As expected, the analysis also showed that CARS was a sensitive indicator of the impact of decision support tools on controller operations. Suggestions for future CARS development and its improvement are also provided.

Lee, Katharine K.↗

NASA NTRS1995

Experimental evaluation of dynamic data allocation strategies in a distributed database with changing workloads

Traditionally, allocation of data in distributed database management systems has been determined by off-line analysis and optimization. This technique works well for static database access patterns, but is often inadequate for frequently changing workloads. In this paper we address how to dynamically reallocate data for partionable distributed databases with changing access patterns. Rather than complicated and expensive optimization algorithms, a simple heuristic is presented and shown, via an implementation study, to improve system throughput by 30 percent in a local area network based system. Based on artificial wide area network delays, we show that dynamic reallocation can improve system throughput by a factor of two and a half for wide area networks. We also show that individual site load must be taken into consideration when reallocating data, and provide a simple policy that incorporates load in the reallocation decision.

Brunstrom, Anna↗

NASA NTRS1979

NASTRAN processing at Johnson Space Center

The processing behavior of NASTRAN, a general purpose system of computer software used for solving engineering analysis problems by means of the finite element displacement method, is discussed. The host system for the NASTRAN is the Univac 1110, and the primary function is performing static and dynamic structural analyses of the Space Shuttle. The processing implications of the NASTRAN structural analysis are addressed and the current NASTRAN situation on the Univac 1110 is examined. The current NASTRAN workload is described and the effect of workload growth due to increased structure size is estimated. A cost and performance analysis is presented.

Richards, F. M., Jr.↗

NASA NTRS1994

Dependability analysis of parallel systems using a simulation-based approach

The analysis of dependability in large, complex, parallel systems executing real applications or workloads is examined in this thesis. To effectively demonstrate the wide range of dependability problems that can be analyzed through simulation, the analysis of three case studies is presented. For each case, the organization of the simulation model used is outlined, and the results from simulated fault injection experiments are explained, showing the usefulness of this method in dependability modeling of large parallel systems. The simulation models are constructed using DEPEND and C++. Where possible, methods to increase dependability are derived from the experimental results. Another interesting facet of all three cases is the presence of some kind of workload of application executing in the simulation while faults are injected. This provides a completely new dimension to this type of study, not possible to model accurately with analytical approaches.

Sawyer, Darren Charles↗

DOE OSTI2021

Evaluating MPI resource usage summary statistics

The Message Passing Interface (MPI) remains the dominant programming model for scientific applications running on today’s high-performance computing (HPC) systems. This dominance stems from MPI’s powerful semantics for inter-process communication that has enabled scientists to write applications for simulating important physical phenomena. MPI does not, however, specify how messages and synchronization should be carried out. Those details are typically dependent on low-level architecture details and the message characteristics of the application. Therefore, analyzing an application’s MPI resource usage is critical to tuning MPI’s performance on a particular platform. The result of this analysis is typically a discussion of the mean message sizes, queue search lengths and message arrival times for a workload or set of workloads. While a discussion of the arithmetic mean in MPI resource usage might be the most intuitive summary statistic, it is not always the most accurate in terms of representing the underlying data. In this paper, we analyze MPI resource usage for a number of key MPI workloads using an existing MPI trace collector and discrete-event simulator. Our analysis demonstrates that the average, while easy and efficient to calculate, is a useful metric for characterizing latency and bandwidth measurements, but may not be a good representation of application message sizes, match list search depths, or MPI inter-operation times. Additionally, we show that the median and mode are superior choices in many cases. We also observe that the arithmetic mean is not the best representation of central tendency for data that are drawn from distributions that are multi-modal or have heavy tails. Furthermore, the results and analysis of our work provide valuable guidance on how we, as a community, should discuss and analyze MPI resource usage data for scientific applications.

97 MATHEMATICS AND COMPUTING↗

DOE OSTI2024

PanDA: Production and Distributed Analysis System

The Production and Distributed Analysis (PanDA) system is a data-driven workload management system engineered to operate at the LHC data processing scale. The PanDA system provides a solution for scientific experiments to fully leverage their distributed heterogeneous resources, showcasing scalability, usability, flexibility, and robustness. The system has successfully proven itself through nearly two decades of steady operation in the ATLAS experiment, addressing the intricate requirements such as diverse resources distributed worldwide at about 200 sites, thousands of scientists analyzing the data remotely, the volume of processed data beyond the exabyte scale, dozens of scientific applications to support, and data processing over several billion hours of computing usage per year. PanDA’s flexibility and scalability make it suitable for the High Energy Physics community and wider science domains at the Exascale. Beyond High Energy Physics, PanDA’s relevance extends to other big data sciences, as evidenced by its adoption in the Vera C. Rubin Observatory and the sPHENIX experiment. As the significance of advanced workflows continues to grow, PanDA has transformed into a comprehensive ecosystem, effectively tackling challenges associated with emerging workflows and evolving computing technologies. The paper discusses PanDA’s prominent role in the scientific landscape, detailing its architecture, functionality, deployment strategies, project management approaches, results, and evolution into an ecosystem.

97 MATHEMATICS AND COMPUTING↗

NASA NTRS1981

A study on task difficulty and acceleration stress

The results of two experiments which relate to task difficulty and the effects of environmental stress on tracking performance are discussed and compared to subjective evaluations. The first experiment involved five different sum of sine tracking tasks which humans tracked both in a static condition and under a 5 Gz acceleration stress condition. The second experiment involved similar environmental stress conditions but in this case the tasks were constructed from deterministic functions with specially designed velocity and acceleration profiles. Phase Plane performance analysis was conducted to study potential measures of workload or tracking difficulty.

Repperger, D. W.↗

NASA NTRS1986

Data communication requirements for the advanced NAS network

The goal of the Numerical Aerodynamic Simulation (NAS) Program is to provide a powerful computational environment for advanced research and development in aeronautics and related disciplines. The present NAS system consists of a Cray 2 supercomputer connected by a data network to a large mass storage system, to sophisticated local graphics workstations, and by remote communications to researchers throughout the United States. The program plan is to continue acquiring the most powerful supercomputers as they become available. In the 1987/1988 time period it is anticipated that a computer with 4 times the processing speed of a Cray 2 will be obtained and by 1990 an additional supercomputer with 16 times the speed of the Cray 2. The implications of this 20-fold increase in processing power on the data communications requirements are described. The analysis was based on models of the projected workload and system architecture. The results are presented together with the estimates of their sensitivity to assumptions inherent in the models.

Levin, Eugene↗

NASA NTRS1992

Design and evaluation of an advanced air-ground data-link system for air traffic control

The design and evaluation of the ground-based portion of an air-ground data-link system for air traffic control (ATC) are described. The system was developed to support the 4D Aircraft/ATC Integration Study, a joint simulation experiment conducted at NASA's Ames and Langley Research Centers. The experiment focused on airborne and ground-based procedures for handling aircraft equipped with a 4D-Flight Management System (FMS) and the system requirements needed to ensure conflict-free traffic flow. The Center/TRACON Automation System (CTAS) at Ames was used for the ATC part of the experiment, and the 4D-FMS-equipped aircraft was simulated by the Transport Systems Research Vehicle (TSRV) simulator at Langley. The data-link system supported not only conventional ATC communications, but also the communications needed to accommodate the 4D-FMS capabilities of advanced aircraft. Of great significance was the synergism gained from integrating the data link with CTAS. Information transmitted via the data link was used to improve the monitoring and analysis capability of CTAS without increasing controller input workload. Conversely, CTAS was used to anticipate and create prototype messages, thus reducing the workload associated with the manual creation of data-link messages.

Denbraven, Wim↗

NASA NTRS2003

Job Superscheduler Architecture and Performance in Computational Grid Environments

Computational grids hold great promise in utilizing geographically separated heterogeneous resources to solve large-scale complex scientific problems. However, a number of major technical hurdles, including distributed resource management and effective job scheduling, stand in the way of realizing these gains. In this paper, we propose a novel grid superscheduler architecture and three distributed job migration algorithms. We also model the critical interaction between the superscheduler and autonomous local schedulers. Extensive performance comparisons with ideal, central, and local schemes using real workloads from leading computational centers are conducted in a simulation environment. Additionally, synthetic workloads are used to perform a detailed sensitivity analysis of our superscheduler. Several key metrics demonstrate that substantial performance gains can be achieved via smart superscheduling in distributed computational grids.

Shan, Hongzhang↗

NASA NTRS2001

Prospective memory in dynamic environments: effects of load, delay, and phonological rehearsal

A new paradigm was developed to examine prospective memory performance in a visual-spatial task that resembles some aspects of the work of air traffic controllers. Two experiments examined the role of workload (number of aeroplanes that participants directed), delay (between receipt of prospective instructions and execution), and phonological rehearsal. High workload increased prospective memory errors but increasing delay from 1-3 or 5 minutes had no effect. Shadowing aurally presented text reduced prospective memory performance, presumably because it prevented verbal rehearsal of the prospective instructions. However, performance on the foreground task of directing aeroplanes to routine destinations was affected only by workload and not by opportunity for rehearsal. Our results suggest that ability to maintain performance on a routine foreground task while performing a secondary task--perhaps analogous to conversation--does not predict ability to retrieve a prospective intention to deviate from the routine.

Aerospace Medicine↗

NASA NTRS1988

Identification of visual evoked response parameters sensitive to pilot mental state

Systems analysis techniques were developed and demonstrated for modeling the electroencephalographic (EEG) steady state visual evoked response (ssVER), for use in EEG data compression and as an indicator of mental workload. The study focused on steady state frequency domain stimulation and response analysis, implemented with a sum-of-sines (SOS) stimulus generator and an off-line describing function response analyzer. Three major tasks were conducted: (1) VER related systems identification material was reviewed; (2) Software for experiment control and data analysis was developed and implemented; and (3) ssVER identification and modeling was demonstrated, via a mental loading experiment. It was found that a systems approach to ssVER functional modeling can serve as the basis for eventual development of a mental workload indicator. The review showed how transient visual evoked response (tVER) and ssVER research are related at the functional level, the software development showed how systems techniques can be used for ssVER characterization, and the pilot experiment showed how a simple model can be used to capture the basic dynamic response of the ssVER, under varying loads.

Zacharias, G. L.↗

NASA NTRS1988

Voice-stress measure of mental workload

In a planned experiment, male subjects between the age of 18 and 50 will be required to produce speech while performing various tasks. Analysis of the speech produced should reveal which aspects of voice prosody are associated with increased workloads. Preliminary results with two female subjects suggest a possible trend for voice frequency and amplitude to be higher and the variance of the voice frequency to be lower in the high workload condition.

Alpert, Murray↗

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↗

NASA NTRS1985

Effect of system workload on operating system reliability - A study on IBM 3081

This paper presents an analysis of operating system failures on an IBM 3081 running VM/SP. Three broad categories of software failures are found: error handling, program control or logic, and hardware related; it is found that more than 25 percent of software failures occur in the hardware/software interface. Measurements show that results on software reliability cannot be considered representative unless the system workload is taken into account. The overall CPU execution rate, although measured to be close to 100 percent most of the time, is not found to correlate strongly with the occurrence of failures. Possible reasons for the observed workload failure dependency, based on detailed investigations of the failure data, are discussed.

Iyer, R. K.↗

NASA NTRS1987

In-flight assessment of workload using instrument scan

During instrument flight, the pilot obtains information concerning aircraft state by cross checking or scanning the flight instruments. The exact method of scanning the instrument panel varies from pilot to pilot but there are some basic features common to a good scan pattern. The method discussed may be considered a candidate for workload studies with piloting tasks which will invoke a regular visual scan (spatial/temporal pattern of eye movements) during instrument flight. It is important to point out that instrument scan by itself is not a complete indicator of workload nor is task attention necessarily associated with where the pilot happens to be looking at a particular instant. However, whenever instrument scan is required in a piloting task, analysis of scanning behavior may yield important direct or indirect information concerning workload.

Tole, J. R.↗