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 559 records · Page 31

Modeling Cooperative Threads to Project GPU Performance for Adaptive Parallelism

Most accelerators, such as graphics processing units (GPUs) and vector processors, are particularly suitable for accelerating massively parallel workloads. On the other hand, conventional workloads are developed for multi-core parallelism, which often scale to only a few dozen OpenMP threads. When hardware threads significantly outnumber the degree of parallelism in the outer loop, programmers are challenged with efficient hardware utilization. A common solution is to further exploit the parallelism hidden deep in the code structure. Such parallelism is less structured: parallel and sequential loops may be imperfectly nested within each other, neigh boring inner loops may exhibit different concurrency patterns (e.g. Reduction vs. Forall), yet have to be parallelized in the same parallel section. Many input-dependent transformations have to be explored. A programmer often employs a larger group of hardware threads to cooperatively walk through a smaller outer loop partition and adaptively exploit any encountered parallelism. This process is time-consuming and error-prone, yet the risk of gaining little or no performance remains high for such workloads. To reduce risk and guide implementation, we propose a technique to model workloads with limited parallelism that can automatically explore and evaluate transformations involving cooperative threads. Eventually, our framework projects the best achievable performance and the most promising transformations without implementing GPU code or using physical hardware. We envision our technique to be integrated into future compilers or optimization frameworks for autotuning.

Meng, Jiayuan↗

Co-design of Advanced Architectures for Graph Analytics using Machine Learning

A graph is an excellent way of representing relationships among entities. We can use graph analytics to synthesize and analyze such relational data, and extract relevant features that are useful for various tasks such as machine learning. Considering the crucial role of graph analytics in various domains, it is important and timely to investigate the right hardware configurations that can achieve optimal performance for graph workloads on future high-performance computing systems. Design space exploration studies facilitate the selection of appropriate configurations (e.g. memory) to achieve a desired system performance. Recently, the approach of accelerating graph analytics using persistent non-volatile memory has gained a lot of attention. Traditional system simulators such as Gem5 and NVMain can be used to explore the design space of these advanced memory architectures for graph workloads. However, these simulators are slow in execution thus limiting the efficiency of design space exploration studies. To overcome this challenge, we proposed a machine learning based approach to co-design advanced memory architectures for graph workloads. We tested our approach with DRAM, non-volatile memory, and hybrid memory (DRAM+NVM) using a breadth first search benchmark algorithm. Our results showed the applicability of the proposed machine learning based approach to the co-design of the advanced memory architectures. In this paper, we provide recommendations on selecting advanced memory architectures to achieve desired performance for graph workloads. We also discuss the performances of different machine learning models that were considered in this study.

Kurte, Kuldeep↗

Dynamic Data Repartitioning for Load-Balanced Parallel Particle Tracing

We present a novel dynamic load-balancing algorithm based on data repartitioning for parallel particle tracing in flow visualization. Instead of static data assignment, we dynamically repartition the data into blocks and reassign the blocks to processes to balance the workload distribution among the processes. Block repartitioning is performed based on a dynamic workload estimation method that predicts the workload in the flow field on the fly as the input. In our approach, we allow data duplication in the repartitioning, enabling the same data blocks to be assigned to multiple processes. Load balance is achieved by regularly exchanging the blocks (together with the particles in the blocks) among processes according to the output of the data repartitioning. Compared with other load-balancing algorithms, our approach does not need any preprocessing on the raw data and does not require any dedicated process for work scheduling, while it has the capability to balance uneven workload efficiently. Results show improved load balance and high efficiency of our method on tracing particles in both steady and unsteady flow.

data repartitioning↗

Preemptible I/O Scheduling of Garbage Collection for Solid State Drives

Unlike hard disks, flash devices use out-of-place updates operations and require a garbage collection (GC) process to reclaim invalid pages to create free blocks. This GC process is a major cause of performance degradation when running concurrently with other I/O operations as internal bandwidth is consumed to reclaim these invalid pages. The invocation of the GC process is generally governed by a low watermark on free blocks and other internal device metrics that different workloads meet at different intervals. This results in an I/O performance that is highly dependent on workload characteristics. In this paper, we examine the GC process and propose a semipreemptible GC (PGC) scheme that allows GC processing to be preempted while pending I/O requests in the queue are serviced. Moreover, we further enhance flash performance by pipelining internal GC operations and merge them with pending I/O requests whenever possible. Our experimental evaluation of this semi-PGC scheme with realistic workloads demonstrates both improved performance and reduced performance variability. Write-dominant workloads show up to a 66.56% improvement in average response time with a 83.30% reduced variance in response time compared to the non-PGC scheme. In addition, we explore opportunities of a new NAND flash device that supports suspend/resume commands for read, write, and erase operations for fully PGC (F-PGC). Our experiments with an F-PGC enabled flash device show that request response time can be improved by up to 14.57% compared to semi-PGC.

97 MATHEMATICS AND COMPUTING↗

Dynamic Undervolting to Improve Energy Efficiency on Multicore X86 CPUs

Chip manufacturers introduce redundancy at various levels of CPU design to guarantee correct operation, even for worst-case combinations of non-idealities in process variation and system operating conditions. This redundancy is implemented partly in the form of voltage margins. However, for a wide range of real-world execution scenarios these margins are excessive and merely translate to increased power consumption, hindering the effort towards higher-energy efficiency in both HPC and general purpose computing. Our study on the x86-64 Haswell and Skylake multicore microarchitectures reveals-wide voltage margins, which vary across different microarchitectures, different chip parts of the same microarchitecture, and across different workloads. We find that it is necessary to quantify-voltage margins using multi-threaded and multi-instance workloads, as characterization with single-threaded and single-instance workloads that do not stress the CPU to its full capacity typically identifies overly optimistic margins that lead to errors when applied in realistic program execution scenarios. In addition, we introduce, deploy and evaluate a run-time governor that dynamically reduces the supply voltage of modern multicore x86-64 CPUs. Our governor employs a model that takes as input a set of performance metrics which are directly measurable via performance monitoring counters and have high predictive value for the minimum tolerable supply voltage (V min ), to predict and apply the appropriate reduction for the workload at hand. Compared with the conventional DVFS governor, our approach in this study achieves up to 42 percent energy savings for the Skylake family and 34 percent for the Haswell family for complex, real-world applications.

97 MATHEMATICS AND COMPUTING↗

DRAS: Deep Reinforcement Learning for Cluster Scheduling in High Performance Computing

Cluster schedulers are crucial in high-performance computing (HPC). They determine when and which user jobs should be allocated to available system resources. Existing cluster scheduling heuristics are developed by human experts based on their experience with specific HPC systems and workloads. However, the increasing complexity of computing systems and the highly dynamic nature of application workloads have placed tremendous burden on manually designed and tuned scheduling heuristics. More aggressive optimization and automation are needed for cluster scheduling in HPC. In this work, we present an automated HPC scheduling agent named DRAS (Deep Reinforcement Agent for Scheduling) by leveraging deep reinforcement learning. DRAS is built on a hierarchical neural network incorporating special HPC scheduling features such as resource reservation and backfilling. An efficient training strategy is presented to enable DRAS to rapidly learn the target environment. Once being provided a specific scheduling objective given by the system manager, DRAS automatically learns to improve its policy through interaction with the scheduling environment and dynamically adjusts its policy as workload changes. We implement DRAS into a HPC scheduling platform called CQGym. CQGym provides a common platform allowing users to flexibly evaluate DRAS and other scheduling methods such as heuristic and optimization methods. Furthermore, the experiments using CQGym with different production workloads demonstrate that DRAS outperforms the existing heuristic and optimization approaches by up to 50%.

97 MATHEMATICS AND COMPUTING↗

Scatter radiation intensities around a clinical digital breast tomosynthesis unit and the impact on radiation shielding considerations

Purpose: To measure the scattered radiation intensity around a clinical digital breast tomosynthesis (DBT) unit and to provide updated data for radiation shielding design for DBT systems with tungsten-anode x-ray tubes. Methods: The continuous distribution of scattered x-rays from a clinical DBT system (Hologic Selenia Dimensions) was measured within an angular range of 0°–180° using a linear-array x-ray detector (X-Scan 0.8f3-512, Detection Technology, Inc., Finland), which was calibrated for the x-ray spectrum range of the DBT unit. The effects of x-ray field size, phantom size, and x-ray kVp/filter combination were investigated. Following a previously developed methodology by Simpkin, scatter fraction was determined for the DBT system as a function of angle around the phantom center. Detailed calculations of the scatter intensity from a DBT system were demonstrated using the measured scatter fraction data. Results: For the 30 and 35 kVp acquisition, the scatter-to-primary-ratio and scatter fraction data closely matched with data previously measured by Simpkin. However, the measured data from this study demonstrated the nonisotropic distribution of the scattered radiation around a DBT system, with two strong peaks around 25° and 160°. The majority scatter radiation (>70%) originated from the imaging detector assembly, instead of the phantom. With a workload from a previous survey performed at MGH, the scatter air kerma at 1 m from the phantom center for wall/door is 1.76 × 10{sup −2} mGy patient{sup −1}, for floor is 1.64 × 10{sup −1} mGy patient{sup −1}, and for ceiling is 3.66 × 10{sup −2} mGy patient{sup −1}. Conclusions: Comparing to previously measured data for mammographic systems, the scatter air kerma from Holgoic DBT is at least two times higher. The main reasons include the harder primary beam with higher workload (measured with total mAs/week), added tomosynthesis acquisition, and strong small angle forward scattering. Due to the highly conservative initial assumptions, the shielding recommendation from NCRP Report 147 is still sufficient for the Hologic DBT system given the workload from a previous survey at MGH. With the data provided from this study, accurate shielding calculation can be performed for Hologic DBT systems with specific workload and barrier distance.

61 RADIATION PROTECTION AND DOSIMETRY↗

WE-DE-207B-09: Scatter Radiation Measurement From a Digital Breast Tomosynthesis System and Its Impact On Shielding Consideration

Purpose: To accurately measure the scatter radiation from a Hologic digital breast tomosynthesis (DBT) system and to provide updated scatter distribution to guide radiation shielding calculation for DBT rooms. Methods: A high sensitivity GOS-based linear detector was used to measure the angular distribution of scatter radiation from a Hologic Selenia Dimensions DBT system. The linear detector was calibrated for its energy response of typical DBT spectra. Following the NCRP147 approach, the measured scatter intensity was normalized by the primary beam area and primary air kerma at 1m from the scatter phantom center and presented as the scatter fraction. Direct comparison was made against Simpkin’s initial measurement. Key parameters including the phantom size, primary beam area, and kV/anode/target combination were also studied. Results: The measured scatter-to-primary-ratio and scatter fraction data closely matched with previous data from Simpkin. The measured data demonstrated the unique nonisotropic distribution of the scattered radiation around a Hologic DBT system, with two strong peaks around 25° and 160°. The majority scatter radiation (>70%) originated from the imaging detector assembly, instead of the phantom. With a workload from a previous local survey, the scatter air kerma at 1m from the phantom center for wall/door is 0.018mGy/patient, for floor is 0.164mGy/patient, and for ceiling is 0.037mGy/patient. Conclusion: Comparing to Simpkin’s previous data, the scatter air kerma from Holgoic DBT is at least two times higher. The main reasons include the harder primary beam with higher workload, added tomosynthesis acquisition, and strong small angle forward scattering. Due to the highly conservative initial assumptions, the shielding recommendation from NCRP147 is still sufficient for the Hologic DBT system given the workload from a previous local survey. With the data provided from this study, accurate shielding calculation can be performed for Hologic DBT systems with specific workload and barrier distance.

60 APPLIED LIFE SCIENCES↗

Distributed caching for processing raw arrays

As applications continue to generate multi-dimensional data at exponentially increasing rates, fast analytics to extract meaningful results is becoming extremely important. The database community has developed array databases that alleviate this problem through a series of techniques. In-situ mechanisms provide direct access to raw data in the original format---without loading and partitioning. Parallel processing scales to the largest datasets. In-memory caching reduces latency when the same data are accessed across a workload of queries. However, we are not aware of any work on distributed caching of multi-dimensional raw arrays. In this paper, we introduce a distributed framework for cost-based caching of multi-dimensional arrays in native format. Given a set of files that contain portions of an array and an online query workload, the framework computes an effective caching plan in two stages. First, the plan identifies the cells to be cached locally from each of the input files by continuously refining an evolving R-tree index. In the second stage, an optimal assignment of cells to nodes that collocates dependent cells in order to minimize the overall data transfer is determined. We design cache eviction and placement heuristic algorithms that consider the historical query workload. A thorough experimental evaluation over two real datasets in three file formats confirms the superiority -- by as much as two orders of magnitude -- of the proposed framework over existing techniques in terms of cache overhead and workload execution time.

Zhao, Weijie↗

Software Defined Architectures for Data Analytics

Data analytics applications increasingly are complex workflows composed of phases with very different program behaviors (e.g., graph algorithms and machine learning, algorithms operating on sparse an dense data structures, etc). To reach the levels of efficiency required to process these workflows in real time, upcoming architectures will need to leverage even more workload specialization. If, at one end, we may find even more heterogenous processors composed by a myriad of specialized processing elements, at the other end we may see novel reconfigurable architectures, composed of sets of functional units and memories interconnected with (re)configurable on-chip networks, able to adapt dynamically to adapt the workload characteristics. Field Programmable Gate Arrays are more and more used for accelerating various workloads and, in particular, inferencing in machine learning, providing higher efficiency than other solutions. However, their fine grained nature still leads to issues for the design software and still makes dynamic reconfiguration impractical. Future, more coarse-grained architectures could offer the features to execute diverse workloads at high efficiency while providing better reconfiguration mechanisms for dynamic adaptability. Nevertheless, we argue that the challenges for reconfigurable computing remain in the software. In this position paper, we describe a possible toolchain for reconfigurable architectures targeted at data analytics.

FPGA, Dataflow, Reconfigurable Architectures, Soft↗

Deffe: A Data-efficient Framework for Performance Characterization in Domain-Specific Computing

Predicting workload performance is a crucial task for many architecture and system research. In this paper we present De￿e, a framework to estimate the workload performances under varying architectural configurations. The infrastructure component of Deffe is based on scalable and easy-to-use open-source software components. By casting the performance modeling as transfer learning tasks, the modeling component of Deffe can leverage the learned knowledge on one workload, and “transfer” it to a new workload. Extensive experimental results show that the method can achieve superior testing accuracy with an effective reduction of 32-80x in terms of the amount of required training data.

Liu, Frank↗

Detection of system failures in multi-axes tasks

The effects of the pilot's participation mode in the control task on his workload level and failure detection performance were examined considering a low visibility landing approach. It is found that the participation mode had a strong effect on the pilot's workload, the induced workload being lowest when the pilot acted as a monitoring element during a coupled approach and highest when the pilot was an active element in the control loop. The effects of workload and participation mode on failure detection were separated. The participation mode was shown to have a dominant effect on the failure detection performance, with a failure in a monitored (coupled) axis being detected significantly faster than a comparable failure in a manually controlled axis.

Ephrath, A. R.↗

Adaptation of time line analysis program to single pilot instrument flight research

A data base was developed for SPIFR operation and the program was run. The outputs indicated that further work was necessary on the workload models. In particular, the workload model for the cognitive channel should be modified as the output workload appears to be too small. Included in the needed refinements are models to show the workload when in turbulence, when overshooting a radial or glideslope, and when copying air traffic control clearances.

Hinton, D. A.↗

Practical guidance for the design of controls and displays for single pilot IFR

This paper represents a first step in developing the criteria for pilot interaction with advanced controls and displays in a single pilot IFR (SPIFR) environment. The research program presented is comprised of an analytical phase and an experimental phase. The analytical phase consisted of a review of fundamental considerations for pilot workload taking into account existing data, and using that data to develop a SPIFR pilot workload model. The rationale behind developing such a model was based on the concept that it is necessary to identify and quantify the most important components of pilot workload to guide the experimental phase of the research which consisted of an abbreviated flight test program. The purpose of the flight tests was to evaluate the workload associated with certain combinations of controls and displays in a flight environment. This was accommplished as a first step in building a data base for single pilot IFR controls and displays.

Hoh, R. H.↗

Levels of Information Processing in a Fitts law task (LIPFitts)

State-of-the-art flight technology has restructured the task of human operators, decreasing the need for physical and sensory resources, and increasing the quantity of cognitive effort required, changing it qualitatively. Recent technological advances have the most potential for impacting a pilot in two areas: performance and mental workload. In an environment in which timing is critical, additional cognitive processing can cause performance decrements, and increase a pilot's perception of the mental workload involved. The effects of stimulus processing demands on motor response performance and subjective mental workload are examined, using different combinations of response selection and target acquisition tasks. The information processing demands of the response selection were varied (e.g., Sternberg memory set tasks, math equations, pattern matching), as was the difficulty of the response execution. Response latency as well as subjective workload ratings varied in accordance with the cognitive complexity of the task. Movement times varied according to the difficulty of the response execution task. Implications in terms of real-world flight situations are discussed.

Mosier, K. L.↗

A measurement-based study of concurrency in a multiprocessor

A systematic measurement-based methodology for characterizing the amount of concurrency present in a workload, and the effect of concurrency on system performance indices such as cache miss rate and bus activity are developed. Hardware and software instrumentation of an Alliant FX/8 was used to obtain data from a real workload environment. Results show that 35% of the workload is concurrent, with the concurrent periods typically using all available processors. Measurements of periods of change in concurrency show uneven usage of processors during these times. Other system measures, including cache miss rate and processor bus activity, are analyzed with respect to the concurrency measures. Probability of a cache miss is seen to increase with concurrency. The change in cache miss rate is much more sensitive to the fraction of concurrent code in the worklaod than the number of processors active during concurrency. Regression models are developed to quantify the relationships between cache miss rate, bus activity, and the concurrency measures. The model for cache miss rate predicts an increase in the median miss rate value as much as 300% for a 100% increase in concurrency in the workload.

Mcguire, Patrick John↗

Measurement-based analysis of error latency

This paper demonstrates a practical methodology for the study of error latency under a real workload. The method is illustrated with sampled data on the physical memory activity, gathered by hardware instrumentation on a VAX 11/780 during the normal workload cycle of the installation. These data are used to simulate fault occurrence and to reconstruct the error discovery process in the system. The technique provides a means to study the system under different workloads and for multiple days. An approach to determine the percentage of undiscovered errors is also developed and a verification of the entire methodology is performed. This study finds that the mean error latency, in the memory containing the operating system, varies by a factor of 10 to 1 (in hours) between the low and high workloads. It is found that of all errors occurring within a day, 70 percent are detected in the same day, 82 percent within the following day, and 91 percent within the third day. The increase in failure rate due to latency is not so much a function of remaining errors but is dependent on whether or not there is a latent error.

Chillarege, Ram↗

Operational testing of a figure of merit for overall task performance

An overall indicator or figure of merit (FOM), for the quality of pilot performance is needed to define optimal workload levels, predict system failure, measure the impact of new automation in the cockpit, and define the relative contributions of subtasks to overall task performance. A normative FOM was developed based on the calculation of a standard score for each component of a complex task. It reflected some effects, detailed in an earlier study, of the introduction of new data link technology into the cockpit. Since the technique showed promise, further testing was done. A new set of data was obtained using the recently developed Multi-Attribute Task Battery. This is a complex battery consisting of four tasks which can be varied in task demand, and on which performance measures can be obtained. This battery was presented to 12 subjects in a 20 minute trial at each of three levels of workload or task demand, and performance measures collected on all four tasks. The NASA-TLX workload rating scale was presented at minutes 6, 12, and 18, of each trial. A figure of merit was then obtained for each run of the battery by calculating a mean, SD, and standard score for each task. Each task contributed its own proportion to the overall FOM, and relative contributions changed with increasing workload. Thus, the FOM shows the effect of task changes, not only on the individual task that is changed, but also on the performance of other tasks and of the whole task. The cost to other tasks of maintaining constant performance on an individual task can be quantified.

Lemay, Moira↗