An Evaluation of Ethernet Performance for Scientific Workloads.
Abstract not provided.
SEARCH · Search NASA
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.
Abstract not provided.
Nuclear power continues to have an imperative role for the U.S.’s electricity generation. However, for these nuclear power plants (NPPs) to remain economically viable, new strategies for reducing operations and maintenance costs must be explored. The U.S. Department of Energy Light Water Reactor Sustainability Program is researching how to digitally transform existing analog and hybrid main control rooms into a fully integrated control room that addresses this challenge. The transformation will fundamentally change the conduct of operations for these U.S. NPPs. Human factors engineering has a vital role in this effort, where traditional task analysis methods are important in informing the design of advanced human-system interface displays. This work describes a preliminary tool to support task analysis for the development of these advanced displays. Details on the use of this tool, including the specific task analysis methods offered, are presented in this paper.
Abstract not provided.
Abstract not provided.
Explore the source record for details and available documents.
Abstract not provided.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
Explore the source record for details and available documents.
High performance computing (HPC) scheduling landscape currently faces new challenges due to thechanges in the workload. Previously, HPC centers were dominated by tightly coupled MPI jobs. HPCworkloads increasingly include high-throughput, data-intensive, and stream-processing applications. Asa consequence, workloads are becoming more diverse at both application and job levels, posing newchallenges to classical HPC schedulers. There is a need to understand the current HPC workloads andtheir evolution to facilitate informed future scheduling research and enable efficient scheduling in futureHPC systems.In this paper, we present a methodology to characterize workloads and assess their heterogeneity,at a particular time period and its evolution over time. We apply this methodology to the workloads ofthree systems (Hopper, Edison, and Carver) at the National Energy Research Scientific Computing Center(NERSC). We present the resulting characterization of jobs, queues, heterogeneity, and performance thatincludes detailed information of a year of workload (2014) and evolution through the systems’ lifetime(2010–2014).
The purpose of this study was to compare the U.S. and Russian cycle ergometers focusing on the mechanical differences of the devices and the physiological differences observed while using the devices. Methods: First, the mechanical loads provided by the U.S. Cycle Ergometer with Vibration Isolation System (CEVIS) and the Russian Veloergometer were measured using a calibration dynamometer. Results were compared and conversion equations were modeled to determine the actual load provided by each device. Second, ten male subjects (32.9 +/- 6.5 yrs, 180.6 +/- 4.4 cm; 81.9 +/- 6.9 kg) experienced with both cycling and exercise testing completed a standardized submaximal exercise test protocol on CEVIS and Veloergometer. The exercise protocol involved 8 sub-maximal workloads each lasting 3 minutes for a total of 24 minutes per session, or until the end of the stage when the subject reached 85% of peak oxygen consumption or age-predicted maximum heart rate (220 - age). The workload started at 50 Watts (W), increased to 100 W, and then increased 25 W every 3 minutes until reaching a peak workload of 250 W. Physiological variables were then compared at each workload by repeated measures ANOVA or paired t-tests (p<0.05). Results: While both CEVIS and Veloergometer produced significantly lower workloads than the displayed workload, CEVIS produced even lower loads than Veloergometer (p<0.05) at each indicated workload. Despite this fact, the only physiological variables that showed a significant difference between the ergometers were VE (125 - 250W), VO2 (175 and 250 W), and VCO2 (175 W). All other physiological data were not statistically different between CEVIS and Veloergometer. Conclusion: Although workloads were different between ergometers, relatively few physiological differences were observed. Therefore, CEVIS workloads of 87.5 - 262.5 W can be rounded to the nearest 25 W increment and performed on the Veloergometer.
A new cost function representing air traffic controller workload is implemented in the Geosect airspace partitioning tool. Geosect currently uses a combination of aircraft count and dwell time to select optimal airspace partitions that balance controller workload. This is referred to as the aircraft count/dwell time hybrid cost function. The new cost function is based on Simplified Dynamic Density, a measure of different aspects of air traffic controller workload. Three sectorizations are compared. These are the current sectorization, Geosect's sectorization based on the aircraft count/dwell time hybrid cost function, and Geosect s sectorization based on the Simplified Dynamic Density cost function. Each sectorization is evaluated for maximum and average workload along with workload balance using the Simplified Dynamic Density as the workload measure. In addition, the Airspace Concept Evaluation System, a nationwide air traffic simulator, is used to determine the capacity and delay incurred by each sectorization. The sectorization resulting from the Simplified Dynamic Density cost function had a lower maximum workload measure than the other sectorizations, and the sectorization based on the combination of aircraft count and dwell time did a better job of balancing workload and balancing capacity. However, the current sectorization had the lowest average workload, highest sector capacity, and the least system delay.
This paper presents a workload classification framework that discriminates illicit computation from authorized workloads on GPU-accelerated HPC systems. As such systems become more and more powerful, they are exploited by attackers to run malicious and for-profit programs that typically require extremely high computing ability to be successful. Our classification framework leverages the distinctive signatures between illicit and authorized workloads, and explore machine learning methods to learn the workloads and classify them. The framework uses lightweight, non-intrusive workload profiling to collect model input data, and explores multiple machine learning methods, particularly recurrent neural network (RNN) that is suitable for online anomalous workload detection. Evaluation results on four generations of GPU machines demonstrate that the workload classification framework can tell apart the illicit authorized workloads with a high accuracy of over 95%.