Search NASA⌕ Search

SEARCH · Search NASA

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 631 records · Page 35

Spy in the GPU-box: Covert and Side Channel Attacks on Multi-GPU System

The deep learning revolution has been enabled in large part by GPUs, and more recently accelerators, which make it possible to carry out computationally demanding training and inference in acceptable times. As the size of machine learning networks and workloads continues to increase, multi-GPU machines have emerged as an important platform offered on High Performance Computing and cloud data centers. Since these machines are shared among multiple users, it becomes increasingly important to protect applications against potential attacks. In this paper, we explore the vulnerability of Nvidia's DGX multi-GPU machines to covert and side channel attacks. These machines consist of a number of discrete GPUs that are interconnected through a combination of custom interconnect (NVLink) and PCIe connections. We reverse engineer the interconnected cache hierarchy and show that it is possible for an attacker on one GPU to cause contention on the L2 cache of another GPU. We use this observation to first develop a covert channel attack across two GPUs, achieving the best bandwidth of around 4 MB/s. We also develop a prime and probe attack on a remote GPU allowing an attacker to recover the cache access pattern of another workload. This access pattern can be used in any number of side channel attacks: we demonstrate a proof of concept attack that fingerprints the application running on the remote GPU, with high accuracy. We also develop a proof of concept attack to extract hyperparameters of a machine learning workload. Our work establishes for the first time the vulnerability of these machines to microarchitectural attacks and can guide future research to improve their security.

Dutta, Sankha↗

Compiler-Driven FPGA Virtualization with SYNERGY

FPGAs are increasingly common in modern applications, and cloud providers now support on-demand FPGA acceleration in datacenters. Applications in datacenters run on virtual infrastructure, where consolidation, multi-tenancy, and workload migration enable economies of scale that are fundamental to the provider's business. However, a general strategy for virtualizing FPGAs has yet to emerge. While manufacturers struggle with hardware-based approaches, we propose a compiler/runtime-based solution called Synergy. We show a compiler transformation for Verilog programs that produces code able to yield control to software atsub-clock-tickgranularity according to the semantics of the original program. Synergy uses this property to efficiently support core virtualization primitives: suspend and resume, program migration, and spatial/temporal multiplexing, on hardware which is availabletoday.We use Synergy to virtualize FPGA workloads across a cluster of Intel SoCs and Xilinx FPGAs on Amazon F1. The workloads require no modification, run within 3--4x of unvirtualized performance, and incur a modest increase in FPGA fabric usage.

Computer Science↗

Performance Study of CXL Memory Topology

This paper presents a comprehensive evaluation of the performance impact of various Compute Express Link (CXL) memory topologies, with a particular emphasis on CXL switches, in the context of High- Performance Computing (HPC) and Large Language Model (LLM) inference workloads. Our study unveils significant performance variations across different topologies, demonstrating that certain configurations yield superior performance for specific workloads. These findings underscore the critical importance of tailored topol- ogy selection in optimizing system performance. Additionally, we address the inherent challenges associated with integrating CXL switches, including overhead considerations and routing complex- ities. Our research highlights the necessity for thorough evalua- tion methodologies to fully leverage CXL technology’s potential in contemporary computing environments. These insights provide valuable guidance for system architects and data center operators in designing and optimizing CXL-based infrastructures for diverse workload requirements.

CXL, memory, Artificial Intelligence (AI), HPC↗

Characterizing the Impact of GPU Power Management on an Exascale System

As GPU-accelerated high-performance computing (HPC) systems approach exascale performance, controlling energy consumption without compromising throughput is essential. Architectures such as the AMD MI250X-based Frontier supercomputer provide runtime mechanisms like frequency and power capping, enabling energy tuning without modifying application code. Although both target energy reduction, they operate via distinct hardware control paths and influence workloads differently. We present a comprehensive evaluation of these strategies on a leadership-class system using diverse HPC proxy applications representative of production workloads. Our study analyzes performance–energy trade-offs across multiple capping levels, node counts (1 and 32), and application profiles. Results show that frequency capping generally achieves higher energy efficiency and scalability, with gains of up to 13.2% without performance loss, while power capping is more effective for single-node runs or bursty GPU utilization. We also provide practical guidelines to help system administrators and users balance energy efficiency and performance in large-scale scientific workloads.

Costa, Mariana [Universidade Federal do Rio Grande↗

Performance trade-offs in reconfigurable networks for HPC

Designing efficient interconnects to support high-bandwidth and low-latency communication is critical toward realizing high performance computing (HPC) and data center (DC) systems in the exascale era. At extreme computing scales, providing the requisite bandwidth through overprovisioning becomes impractical. These challenges have motivated studies exploring reconfigurable network architectures that can adapt to traffic patterns at runtime using optical circuit switching. Despite the plethora of proposed architectures, surprisingly little is known about the relative performances and trade-offs among different reconfigurable network designs. We aim to bridge this gap by tackling two key issues in reconfigurable network design. First, we study how cost, power consumption, network performance, and scalability vary based on optical circuit switch (OCS) placement in the physical topology. Specifically, we consider two classes of reconfigurable architectures: one that places OCSs between top-of-rack (ToR) switches—ToR-reconfigurable networks (TRNs)—and one that places OCSs between pods of racks—pod-reconfigurable networks (PRNs). Second, we tackle the effects of reconfiguration frequency on network performance. Our results, based on network simulations driven by real HPC and DC workloads, show that while TRNs are optimized for low fan-out communication patterns, they are less suited for carrying high fan-out workloads. PRNs exhibit better overall trade-off, capable of performing comparably to a fully non-blocking fat tree for low fan-out workloads, and significantly outperform TRNs for high fan-out communication patterns.

Teh, Min Yee↗

Stacked filters: learning to filter by structure

We present Stacked Filters, a new probabilistic filter which is fast and robust similar to query-agnostic filters (such as Bloom and Cuckoo filters), and at the same time brings low false positive rates and sizes similar to classifier-based filters (such as Learned Filters). The core idea is that Stacked Filters incorporate workload knowledge about frequently queried non-existing values. Instead of learning, they structurally incorporate that knowledge using hashing and several sequenced filter layers, indexing both data and frequent negatives. Stacked Filters can also gather workload knowledge on-the-fly and adaptively build the filter. We show experimentally that for a given memory budget, Stacked Filters achieve end-to-end query throughput up to 130x better than the best alternative for a workload, either query-agnostic or classifier-based filters, and depending on where data is (SSD or HDD).

Computer Science↗

Evaluating Function-as-a-Service (FaaS) frameworks for the Accelerator Control System

As particle accelerator control systems evolve in complexity and scale, the need for responsive, scalable, and cost-effective computational infrastructure becomes increasingly critical. Function-as-a-Service (FaaS) offers an alternative to traditional monolithic architecture by enabling event-driven execution, automatic scaling, and fine-grained resource utilization. This paper explores the applicability and performance of FaaS frameworks in the context of a modern particle accelerator control system, with the objective of evaluating their suitability for short lived and triggered workloads. In this paper, we evaluate prominent open-source FaaS platforms in executing functional logic, triggers, and diagnostics routines. Evaluation metrics consist of cold-start latency, scalability, performance, integration with other open-source tools like Kafka. Experimental workloads were designed to simulate real-world control tasks when implemented as stateless FaaS functions. These workloads were benchmarked under various invocation loads and network conditions. Self-hosted FaaS platforms, when deployed within accelerator networks, offer greater control over execution environment, better integration with legacy systems, and support for real-time guarantees when paired with message queues. Based on lessons learned and evaluation metrics, this paper describes reliability of the FaaS framework for the Accelerator Control Systems (ACS).

Jaikar, A. [Fermilab] (ORCID:0000000332046217)↗

Performance Evaluation of Darshan 3.0.0 on the Cray XC30

Abstract—Darshan is a lightweight I/O characterization tool used to gather and summarize salient I/O workload statistics from HPC applications. Darshan was designed to minimize any possible perturbations of an application’s performance, leading it to be enabled by default on a number of production HPC systems. For each file accessed by a given application, Darshan records the count and types of I/O operations performed, histograms of access sizes, cumulative timers on the amount of time spent doing I/O, and other statistical data. This type of data has proved invaluable in understanding and improving the I/O performance of HPC applications. Darshan 3.0.0 is the new modularized version of the traditional Darshan library and file format, allowing users to easily add more in-depth I/O characterization data to Darshan logs. In this work we perform an empirical evaluation of Darshan 3.0.0 to ensure that it continues to meet performance expectations for broad deployment. In particular, we evaluate the imposed overhead on instrumented I/O operations, time taken to shut down and generate corresponding Darshan log files, and resultant log file sizes for different workloads. These performance results are compared to results of Darshan 2.3.0 on the Edison XC30 system at NERSC to determine whether the new version is lightweight enough to run full-time on production HPC systems. Our evaluation shows that Darshan has limited impact on application I/O performance and can fully generate a corresponding log file for most application workloads in under two seconds.

97 MATHEMATICS AND COMPUTING↗

Enabling Diverse Software Stacks on Supercomputers using High Performance Virtual Clusters.

While large-scale simulations have been the hallmark of the High Performance Computing (HPC) community for decades, Large Scale Data Analytics (LSDA) workloads are gaining attention within the scientific community not only as a processing component to large HPC simulations, but also as standalone scientific tools for knowledge discovery. With the path towards Exascale, new HPC runtime systems are also emerging in a way that differs from classical distributed com- puting models. However, system software for such capabilities on the latest extreme-scale DOE supercomputing needs to be enhanced to more appropriately support these types of emerging soft- ware ecosystems. In this paper, we propose the use of Virtual Clusters on advanced supercomputing resources to enable systems to support not only HPC workloads, but also emerging big data stacks. Specifi- cally, we have deployed the KVM hypervisor within Cray's Compute Node Linux on a XC-series supercomputer testbed. We also use libvirt and QEMU to manage and provision VMs directly on compute nodes, leveraging Ethernet-over-Aries network emulation. To our knowledge, this is the first known use of KVM on a true MPP supercomputer. We investigate the overhead our solution using HPC benchmarks, both evaluating single-node performance as well as weak scaling of a 32-node virtual cluster. Overall, we find single node performance of our solution using KVM on a Cray is very efficient with near-native performance. However overhead increases by up to 20% as virtual cluster size increases, due to limitations of the Ethernet-over-Aries bridged network. Furthermore, we deploy Apache Spark with large data analysis workloads in a Virtual Cluster, ef- fectively demonstrating how diverse software ecosystems can be supported by High Performance Virtual Clusters.

97 MATHEMATICS AND COMPUTING↗

HyLiPoD: Parallel Particle Advection via a Hybrid of Lifeline Scheduling and Parallelization-Over-Data

Performance characteristics of parallel particle advection algorithms can vary greatly based on workload.With this short paper, we build a new algorithm based on results from a previous bake-off study which evaluated the performance of four algorithms on a variety of workloads. Our algorithm, called HyLiPoD, is a ''meta-algorithm,'' i.e., it considers the desired workload to choose from existing algorithms to maximize performance. To demonstrate HyliPoD's benefit, we analyze results from 162 tests including concurrencies of up to 8192 cores, meshes as large as 34 billion cells, and particle counts as large as 300 million. Our findings demonstrate that HyLiPoD's adaptive approach allows it to match the best performance of existing algorithms across diverse workloads.

high performance computing↗

Quantifying automated vehicle benefits in reducing driving stress: a simulation experiment approach

Driving is a stressful activity because of the mental workload required to maneuver a vehicle in certain travel contexts, such as congested traffic, multi-modal networks requiring complex interaction with surrounding vehicles, and aggressive driving. Autonomous vehicles (AVs), on the other hand, can reduce the mental workload by performing most of the driving tasks and providing users with a comfortable ride. This study develops a pathway model to relate different health determinants, including travel reliability, safety, driving comfort, and value of time, to Autonomous vehicles driving and studies their impact on the value of driving stress. A case study example of Autonomous vehicles simulation is used to determine the impact of these health determinants. The value of driving stress in Autonomous vehicles is estimated as a function of the value of these individual health determinants. The results show that the perception of safe or unsafe driving in Autonomous vehicles is the most important factor in changing the perception of driving stress in Autonomous vehicles. Similarly, perceptions of comfortable driving in Autonomous vehicles and reduced workload with a higher value of time also reduce driving stress in Autonomous vehicles. These results allow Autonomous vehicles adoption models to explicitly consider driving stress reduction as a benefit and can improve understanding of Autonomous vehicles adoption, which may require quantitative analysis of underlying motivating benefits, including driving stress reduction.

Khattak, Zulqarnain H.↗

A model based technique for the design of flight directors

A new technique for designing flight directors is discussed. This technique uses the optimal-control pilot/vehicle model to determine the appropriate control strategy. The dynamics of this control strategy are then incorporated into the director control laws, thereby enabling the pilot to operate at a significantly lower workload. A preliminary design of a control director for maintaining a STOL vehicle on the approach path in the presence of random air turbulence is evaluated. By selecting model parameters in terms of allowable path deviations and pilot workload levels, a set of director laws is achieved which allows improved system performance at reduced workload levels. The pilot acts essentially as a proportional controller with regard to the director signals, and control motions are compatible with those appropriate to status-only displays.

Levison, W. H.↗

Pilot performance in zero-visibility precision approach

The pilot's short-term decisions regarding performance assessment and failure monitoring is examined. The performance of airline pilots who flew simulated zero-visibility landing approaches is reported. Results indicate that the pilot's mode of participation in the control task has a strong effect on his workload, the induced workload being lowest when the pilot acts as a monitor during a coupled approach and highest when the pilot is an active element in the control loop. A marked increase in workload at altitudes below 500 ft. is documented at all participation modes; this increase is inversely related to distance-to-go. The participation mode is shown to have a dominant effect on failure-detection performance, with a failure in a monitored (coupled) axis being detected faster than a comparable failure in a manually-controlled axis. Touchdown performance is also documented. It is concluded that the conventional instrument panel and its associated displays are inadequate for zero-visibility operations in the final phases of the landing approach.

Ephrath, A. R.↗

Manual and automatic flight control during severe turbulence penetration

An analytical and experimental investigation of possible contributing factors in jet aircraft turbulence upsets was conducted. Major contributing factors identified included autopilot and display deficiencies, the large aircraft inertia and associated long response time, and excessive pilot workload. An integrated flight and thrust energy management director system was synthesized. The system was incorporated in a moving-base simulation and evaluated using highly experienced airline pilots. The evaluation included comparison of pilot workload and flight performance during severe turbulence penetration utilizing four control/display concepts: manual control with conventional full panel display, conventional autopilot (A/P-A) with conventional full panel display, improved autopilot (A/P-B) with conventional full panel display plus thrust director display, and longitudinal flight director with conventional full panel display plus thrust director display. Simulation results show improved performance, reduced pilot workload, and a pilot preference for the autopilot system controlling to the flight director command and manual control of thrust following the trim thrust director.

Johnston, D. E.↗

Survey of piloting factors in V/STOL aircraft with implications for flight control system design

Flight control system design factors involved for pilot workload relief are identified. Major contributors to pilot workload include configuration management and control and aircraft stability and response qualities. A digital fly by wire stability augmentation, configuration management, and configuration control system is suggested for reduction of pilot workload during takeoff, hovering, and approach.

Ringland, R. F.↗

Adaptive allocation of decisionmaking responsibility between human and computer in multitask situations

As human and computer come to have overlapping decisionmaking abilities, a dynamic or adaptive allocation of responsibilities may be the best mode of human-computer interaction. It is suggested that the computer serve as a backup decisionmaker, accepting responsibility when human workload becomes excessive and relinquishing responsibility when workload becomes acceptable. A queueing theory formulation of multitask decisionmaking is used and a threshold policy for turning the computer on/off is proposed. This policy minimizes event-waiting cost subject to human workload constraints. An experiment was conducted with a balanced design of several subject runs within a computer-aided multitask flight management situation with different task demand levels. It was found that computer aiding enhanced subsystem performance as well as subjective ratings. The queueing model appears to be an adequate representation of the multitask decisionmaking situation, and to be capable of predicting system performance in terms of average waiting time and server occupancy. Server occupancy was further found to correlate highly with the subjective effort ratings.

Chu, Y.-Y.↗

Flight evaluation of the terminal guidance system

The terminal guidance system (TGS) is avionic equipment which gives guidance along a curved descending flight path to a landing. A Cessna 182 was used as the test aircraft and the TGS was installed and connected to the altimeter, DME, RMI, and gyro compass. Approaches were flown by three different pilots. When the aircraft arrives at the termination point, it is set up on final approach for a landing. The TGS provides guidance for curved descending approaches with guideslopes of 6 deg which required, for experienced pilots, workloads that are approximately the same as for an ILS. The glideslope is difficult to track within 1/2 n.m. of the VOR/DME station. The system permits, for experienced pilots, satisfactory approaches with a turn radius as low as 1/2 n.m. and a glideslope of 6 deg. Turn angles have little relation to pilot workload for curved approaches. Pilot experience is a factor for curved approaches. Pilots with low instrument time have difficulty flying steep approaches with small turn radius. Turbulence increases the pilot workload for curved approaches. The TGS does not correct to a given flight path over the ground nor does it adequately compensate for wind drift.

Sandlin, D. R.↗

A flight investigation of simulated data-link communications during single-pilot IFR flight. Volume 2: Flight evaluations

Key problems in single pilot instrument flight operations are in the management of flight data and the processing of cockpit information during conditions of heavy workload. A flight data console was developed to allow simulation of a digital data link to replace the current voice communications stem used in air traffic control. This is a human factors evaluation of a data link communications system to determine how such a system might reduce cockpit workload, improve flight proficiency, and be accepted by general aviation pilots. The need for a voice channel as backup to a digital link is examined. The evaluations cover both airport terminal area operations and full mission instrument flight. Results show that general aviation pilots operate well with a digital data link communications system. The findings indicate that a data link system for pilot/ATC communications, with a backup voice channel, is well accepted by general aviation pilots and is considered to be safer, more efficient, and result in less workload than the current voice system.

Parker, J. F., Jr.↗