Search NASA⌕ Search

SEARCH · Search NASA

Results for “Parallel processing”

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 127 records · Page 7

Reducing the bulk of the bulk synchronous parallel model

For over two decades the dominant means for enabling portable performance of computational science and engineering applications on parallel processing architectures has been the bulk-synchronous parallel programming (BSP) model. Code developers, motivated by performance considerations to minimize the number of messages transmitted, have typically pursued a strategy of aggregating message data into fewer, larger messages. Emerging and future high-performance architectures, especially those seen as targeting Exascale capabilities, provide motivation and capabilities for revisiting this approach. In this paper we explore alternative configurations within the context of a large-scale complex multi-physics application and a proxy that represents its behavior, presenting results that demonstrate some important advantages as the number of processors increases in scale.

97 MATHEMATICS AND COMPUTING↗

A brief parallel I/O tutorial.

This document provides common best practices for the efficient utilization of parallel file systems for analysts and application developers. A multi-program, parallel supercomputer is able to provide effective compute power by aggregating a host of lower-power processors using a network. The idea, in general, is that one either constructs the application to distribute parts to the different nodes and processors available and then collects the result (a parallel application), or one launches a large number of small jobs, each doing similar work on different subsets (a campaign). The I/O system on these machines is usually implemented as a tightly-coupled, parallel application itself. It is providing the concept of a 'file' to the host applications. The 'file' is an addressable store of bytes and that address space is global in nature. In essence, it is providing a global address space. Beyond the simple reality that the I/O system is normally composed of a small, less capable, collection of hardware, that concept of a global address space will cause problems if not very carefully utilized. How much of a problem and the ways in which those problems manifest will be different, but that it is problem prone has been well established. Worse, the file system is a shared resource on the machine - a system service. What an application does when it uses the file system impacts all users. It is not the case that some portion of the available resource is reserved. Instead, the I/O system responds to requests by scheduling and queuing based on instantaneous demand. Using the system well contributes to the overall throughput on the machine. From a solely self-centered perspective, using it well reduces the time that the application or campaign is subject to impact by others. The developer's goal should be to accomplish I/O in a way that minimizes interaction with the I/O system, maximizes the amount of data moved per call, and provides the I/O system the most information about the I/O transfer per request.

99 GENERAL AND MISCELLANEOUS↗

Microstructural characterization and comparison of Ti-6Al-4V manufactured with different additive manufacturing processes

In this work, the microstructures of Ti-6Al-4V manufactured by different additive manufacturing (AM) processes have been characterized and compared. The microstructural features that were characterized are the α lath thickness, grain boundary α (GB-α) thickness, prior β grain size and α colony size. In addition, the microhardnesses were also measured and compared. The microstructure of shaped metal deposited (SMD) Ti-6Al-4V material showed the smallest variations in α lath size, whereas the material manufactured with laser metal wire deposition-0 (LMwD-0) showed the largest variation. The prior β grain size was found to be smaller in material manufactured with powder bed fusion (PBF) as compared with corresponding material manufactured with the directed energy deposition (DED) processes. Parallel bands were only observed in materials manufactured with DED processes while being non-present in material manufactured with PBF processes.

36 MATERIALS SCIENCE↗

Balancing computation and communication power in power constrained clusters

Systems, apparatuses, and methods for balancing computation and communication power in power constrained environments. A data processing cluster with a plurality of compute nodes may perform parallel processing of a workload in a power constrained environment. Nodes that finish tasks early may be power-gated based on one or more conditions. In some scenarios, a node may predict a wait duration and go into a reduced power consumption state if the wait duration is predicted to be greater than a threshold. The power saved by power-gating one or more nodes may be reassigned for use by other nodes. A cluster agent may be configured to reassign the unused power to the active nodes to expedite workload processing.

24 POWER TRANSMISSION AND DISTRIBUTION↗

HEP Software Foundation Community White Paper Working Group - Data Processing Frameworks

Data processing frameworks are an essential part of HEP experiments' software stacks. Frameworks provide a means by which code developers can undertake the essential tasks of physics data processing, accessing relevant inputs and storing their outputs, in a coherent way without needing to know the details of other domains. Frameworks provide essential core services for developers and help deliver a configurable working application to the experiments' production systems. Modern HEP processing frameworks are in the process of adapting to a new computing landscape dominated by parallel processing and heterogeneity, which pose many questions regarding enhanced functionality and scaling that must be faced without compromising the maintainability of the code. In this paper we identify a program of work that can help further clarify the key concepts of frameworks for HEP and then spawn R&D activities that can focus the community's efforts in the most efficient manner to address the challenges of the upcoming experimental program.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗

Performance for GPU exceptions

Techniques for improving performance of accelerated processing devices (“APDs”) when exceptions occur are provided. In APDs, the very large number of parallel processing execution units, and the complexity of the hardware used to execute a large number of work-items in parallel, means that APDs typically stall when an exception occurs (unlike in central processing units (“CPUs”), which are able to execute speculatively and out-of-order). However, the techniques provided herein allow at least some execution to occur past exceptions. Execution past an exception generating instruction occurs by executing instructions that would not lead to a corruption while skipping those that would lead to a corruption. After the exception has been satisfied, execution occurs in a replay mode in which the potentially exception-generating instruction is executed and in which instructions that did not execute in the exception-wait mode are executed. A mask and counter are used to control execution in replay mode.

Gutierrez, Anthony T.↗

Conception and Operation of a New Treatment Center for Radioactive Waste from High Energy Accelerators - 18293

CERN has operated high-energy accelerators for more than 60 years of research in fundamental physics. The side-product of this activity is the generation of low- and intermediate-level radioactive waste, most of which is in solid form and free of surface contamination. Today, one operational year produces about 400 tons of radioactive waste which need to be disposed of towards the final repositories of the two CERN Host States, France and Switzerland. A new facility for the collection, interim storage, characterization and treatment of radioactive waste became operational at CERN in 2015. This facility was specifically conceived for an accelerator complex with a wide range of challenging waste items, which are often unique in their irradiation history as well as in their physical, chemical and radiological characteristics. The disposal rate was ramped up in 2016 with a unique treatment process tailored for metallic low-level waste. In 2017, two parallel processes were successfully tested for the elimination of radioactive cables and for the recycling of valuable metals. This paper gives an overview of the challenges in the design, commissioning and operation of the CERN radioactive waste treatment center. It outlines the methodology followed for the conception of the elimination processes, which can be of guidance for the development of similar facilities. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Performance Evaluation of Different Parallel Programming Models in SCALE-Shift Sequences for Criticality and Shielding Applications [Abstract]

The SCALE code system has been widely used for nuclear criticality safety, reactor physics, radiation shielding, source term generation, and inventory analyses by researchers, industry, and regulatory bodies. Although limited support for shared- and distributed-memory parallel processing was introduced via C++ threading, OpenMP, and MPI, a hybrid parallel programming model with both distributed- and shared-memory parallelism has not been fully supported in the SCALE code system.

Nuclear Criticality Safety Program (NCSP)↗

Traveler: Navigating Task Parallel Traces for Performance Analysis

Understanding the behavior of software in execution is a key step in identifying and fixing performance issues. This is especially important in high performance computing contexts where even minor performance tweaks can translate into large savings in terms of computational resource use. To aid performance analysis, developers may collect an execution trace —a chronological log of program activity during execution. As traces represent the full history, developers can discover a wide array of possibly previously unknown performance issues, making them an important artifact for exploratory performance analysis. However, interactive trace visualization is difficult due to issues of data size and complexity of meaning. Traces represent nanosecond-level events across many parallel processes, meaning the collected data is often large and difficult to explore. The rise of asynchronous task parallel programming paradigms complicates the relation between events and their probable cause. Here, to address these challenges, we conduct a continuing design study in collaboration with high performance computing researchers. We develop diverse and hierarchical ways to navigate and represent execution trace data in support of their trace analysis tasks. Through an iterative design process, we developed Traveler , an integrated visualization platform for task parallel traces. Traveler provides multiple linked interfaces to help navigate trace data from multiple contexts. We evaluate the utility of Traveler through feedback from users and a case study, finding that integrating multiple modes of navigation in our design supported performance analysis tasks and led to the discovery of previously unknown behavior in a distributed array library.

97 MATHEMATICS AND COMPUTING↗

Processing data communications events by awakening threads in parallel active messaging interface of a parallel computer

Processing data communications events in a parallel active messaging interface (`PAMI`) of a parallel computer that includes compute nodes that execute a parallel application, with the PAMI including data communications endpoints, and the endpoints are coupled for data communications through the PAMI and through other data communications resources, including determining by an advance function that there are no actionable data communications events pending for its context, placing by the advance function its thread of execution into a wait state, waiting for a subsequent data communications event for the context; responsive to occurrence of a subsequent data communications event for the context, awakening by the thread from the wait state; and processing by the advance function the subsequent data communications event now pending for the context.

97 MATHEMATICS AND COMPUTING↗

Process–Property–Performance Mapping of Additively Manufactured 316H Stainless Steel Components

The Advanced Materials and Manufacturing Technologies Program is focused on accelerating the development of advanced materials and components fabricated via additive manufacturing, and is using laser powder bed fusion (LPBF) of 316H stainless steel as an initial case study. In the previous fiscal year, miniature high-throughput specimens were printed on multiple LPBF systems to provide initial processing windows to minimize porosity and limit epitaxial grain growth during prints. This fiscal year, scaled builds were completed on three different LPBF systems at ORNL: a GE Concept Laser M2, a Renishaw AM400, and an EOS M290. Builds on the Concept Laser were conducted on multiple powder lots and processing parameter ranges to provide microstructure effects on time-independent and time-dependent mechanical properties. Builds on the Renishaw were produced using Oak Ridge National Laboratory (ORNL)-optimized printing parameters and Argonne National Laboratory (ANL)-optimized printing parameters to compare outcomes of parallel process optimization efforts at different national laboratories on the same LPBF system. Similarly, the build completed on the EOS M290 replicated the processing parameters of builds completed at Los Alamos National Laboratory (LANL). Optical microscopy and electron backscatter diffraction characterization was completed on all builds. In addition to the general round robin characterization, this work-package generated time-independent data, including tensile and fracture toughness test data on scaled Concept Laser builds as a function of processing parameters and post-build heat treatment. This analysis is complimentary to work in parallel work packages aiming to establish heat treatment and processing effects on time-dependent properties. It was found that although the stress-relief heat treatment provides the highest strength at lower-temperatures, tensile strength begins to converge at higher temperatures regardless of heat treatment condition. In addition, the more rigorous solution annealing and hot-isostatic pressing post-build heat treatments result in higher fracture toughness than the stress-relieved condition. The root-causes of the lower fracture toughness of the stress-relieved LPBF 316H material was informed via a stress-relief optimization study on a scaled concept laser print, where it was found that although dislocation recovery was largely complete after only a couple hours at 650°C, the extended hold of the current 24h heat treatment employed on scaled builds likely caused increased carbide volume fractions along the LPBF 316H grain boundaries, thereby deteriorating crack propagation resistance. This trend was seen to become more deleterious with additional increases of stress-relief temperature to 750°C or 850°C. These results have helped inform a new optimal stress-relief annealing condition for LPBF 316H for future campaign testing (650°C for 2h).

36 MATERIALS SCIENCE↗

Process–Property–Performance Mapping of Additively Manufactured 316H Stainless Steel Components

The Advanced Materials and Manufacturing Technologies Program is focused on accelerating the development of advanced materials and components fabricated via additive manufacturing, and is using laser powder bed fusion (LPBF) of 316H stainless steel as an initial case study. In the previous fiscal year, miniature high-throughput specimens were printed on multiple LPBF systems to provide initial processing windows to minimize porosity and limit epitaxial grain growth during prints. This fiscal year, scaled builds were completed on three different LPBF systems at ORNL: a GE Concept Laser M2, a Renishaw AM400, and an EOS M290. Builds on the Concept Laser were conducted on multiple powder lots and processing parameter ranges to provide microstructure effects on time-independent and time-dependent mechanical properties. Builds on the Renishaw were produced using Oak Ridge National Laboratory (ORNL)-optimized printing parameters and Argonne National Laboratory (ANL)-optimized printing parameters to compare outcomes of parallel process optimization efforts at different national laboratories on the same LPBF system. Similarly, the build completed on the EOS M290 replicated the processing parameters of builds completed at Los Alamos National Laboratory (LANL). Optical microscopy and electron backscatter diffraction characterization was completed on all builds. In addition to the general round robin characterization, this work-package generated time-independent data, including tensile and fracture toughness test data on scaled Concept Laser builds as a function of processing parameters and post-build heat treatment. This analysis is complimentary to work in parallel work packages aiming to establish heat treatment and processing effects on time-dependent properties. It was found that although the stress-relief heat treatment provides the highest strength at lower-temperatures, tensile strength begins to converge at higher temperatures regardless of heat treatment condition. In addition, the more rigorous solution annealing and hot-isostatic pressing post-build heat treatments result in higher fracture toughness than the stress-relieved condition. The root-causes of the lower fracture toughness of the stress-relieved LPBF 316H material was informed via a stress-relief optimization study on a scaled concept laser print, where it was found that although dislocation recovery was largely complete after only a couple hours at 650°C, the extended hold of the current 24h heat treatment employed on scaled builds likely caused increased carbide volume fractions along the LPBF 316H grain boundaries, thereby deteriorating crack propagation resistance. This trend was seen to become more deleterious with additional increases of stress-relief temperature to 750°C or 850°C. These results have helped inform a new optimal stress-relief annealing condition for LPBF 316H for future campaign testing (650°C for 2h).

36 MATERIALS SCIENCE↗

Enabling Highly Efficient Capsule Networks Processing Through A PIM-Based Architecture Design

In recent years, the CNNs have achieved great successes in the image processing tasks, e.g., image recognition and object detection. Unfortunately, traditional CNN's classication is found to be easily misled by increasingly complex image features due to the usage of pooling operations, hence unable to preserve accurate position and pose information of the objects. To address this challenge, a novel neural network structure called Capsule Network has been proposed, which introduces equivariance through capsules to signicantly enhance the learning ability for image segmentation and object detection. Due to its requirement of performing a high volume of matrix operations, CapsNets have been generally accelerated on modern GPU platforms that provide highly optimized software library for common deep learning tasks. However, based on our performance characterization on modern GPUs, CapsNets exhibit low effciency due to the special program and execution features of their routing procedure, including massive unshareable intermediate variables and intensive syn- chronizations, which are very dicult to optimize at software level. To address these challenges, we propose a hybrid computing architecture design named PIM-CapsNet. It preserves GPU's on-chip computing capability for accelerating CNN types of layers in CapsNet, while pipelining with an off-chip in-memory acceleration solution that effectively tackles routing procedure's ineffciency by leveraging the processing-in-memory capability of today's 3D stacked memory. Using routing procedure's inherent parallellization feature, our design enables hierarchical improvements on CapsNet inference effciency through minimizing data movement and maximizing parallel processing in memory. Evaluation results demonstrate that our proposed design can achieve substantial improvement on both performance and energy savings for CapsNet inference, with almost zero accuracy loss. The results also suggest good performance scalability in optimizing the routing procedure with increasing network size.

Zhang, Xingyao↗

Skipper-in-CMOS: Nondestructive Readout With Subelectron Noise Performance for Pixel Detectors

The Skipper-in-CMOS image sensor integrates the nondestructive readout capability of skipper charge coupled devices (Skipper-CCDs) with the high conversion gain of a pinned photodiode (PPD) in a CMOS imaging process while taking advantage of in-pixel signal processing. This allows both single photon counting as well as high frame rate readout through highly parallel processing. The first results obtained from a ${15} \times {15}~\mu $ m2 pixel cell of a Skipper-in-CMOS sensor fabricated in Tower Semiconductor’s commercial 180-nm CMOS image sensor process are presented. Measurements confirm the expected reduction of the readout noise with the number of samples down to deep subelectron noise of $0.15\text {e}^ - $ , demonstrating the charge transfer operation from the PPD and the single photon counting operation when the sensor is exposed to light. This article also discusses new testing strategies employed for its operation and characterization.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

pnnl/archive-sprinter

The Archive Sprinter tool is designed to efficiently process synchrophasor measurements from electric power grids and export data signatures that summarize the grid's behavior. Parallel processing will allow data to be processed quickly to enable practical analyses of archives spanning years. The grid's behavior will be summarized using a wide-array of signatures calculated from the input data.

Follum, Jim↗

A Parallel-in-Time Multigrid Approach to Constrained Optimization

Valuable computer simulations such as weather prediction, traffic flow, and resource allocation are time dependent problems. One such example is the viscous Burgers’ equation which can model various systems such as fluids and traffic flow. Further, many of these models require optimization routines to meet the needs of many businesses. Unfortunately, traditional sequential-in-time algorithms have become bottlenecks in solution model generation. This is because computer architectures have begun to rely more and more on concurrency (parallel processing) as opposed to processor power to provide greater performance. To avoid this bottleneck, many researchers use parallelization methods to take advantage of the increasing number of processors in computers. This approach allows for variables in a problem to be solved for simultaneously. As such, parallelization is a crucial requirement to achieve high-performance computing.

97 MATHEMATICS AND COMPUTING↗

SU-E-T-278: Realization of Dose Verification Tool for IMRT Plan Based On DPM

Purpose: To build a Monte Carlo dose verification tool for IMRT Plan by implementing a irradiation source model into DPM code. Extend the ability of DPM to calculate any incident angles and irregular-inhomogeneous fields. Methods: With the virtual source and the energy spectrum which unfolded from the accelerator measurement data,combined with optimized intensity maps to calculate the dose distribution of the irradiation irregular-inhomogeneous field. The irradiation source model of accelerator was substituted by a grid-based surface source. The contour and the intensity distribution of the surface source were optimized by ARTS (Accurate/Advanced Radiotherapy System) optimization module based on the tumor configuration. The weight of the emitter was decided by the grid intensity. The direction of the emitter was decided by the combination of the virtual source and the emitter emitting position. The photon energy spectrum unfolded from the accelerator measurement data was adjusted by compensating the contaminated electron source. For verification, measured data and realistic clinical IMRT plan were compared with DPM dose calculation. Results: The regular field was verified by comparing with the measured data. It was illustrated that the differences were acceptable (<2% inside the field, 2–3mm in the penumbra). The dose calculation of irregular field by DPM simulation was also compared with that of FSPB (Finite Size Pencil Beam) and the passing rate of gamma analysis was 95.1% for peripheral lung cancer. The regular field and the irregular rotational field were all within the range of permitting error. The computing time of regular fields were less than 2h, and the test of peripheral lung cancer was 160min. Through parallel processing, the adapted DPM could complete the calculation of IMRT plan within half an hour. Conclusion: The adapted parallelized DPM code with irradiation source model is faster than classic Monte Carlo codes. Its computational accuracy and speed satisfy the clinical requirement, and it is expectable to be a Monte Carlo dose verification tool for IMRT Plan. Strategic Priority Research Program of the China Academy of Science(XDA03040000); National Natural Science Foundation of China (81101132)

60 APPLIED LIFE SCIENCES↗

Adaptable Technology Integration for Enhanced Worker Safety and Efficiency within the SPD Gloveboxes [Slides]

Surplus Plutonium Disposition (SPD) Overview: (1) Application Background – Location: Savannah River Site, Aiken, South Carolina – High backlog of plutonium material to process – Parallel lines of gloveboxes – Highly manual operations requiring full PPE and reader/worker steps; (2) Development Scope – Development of automated processes and opportunities to provide robust engineering design alternatives for baseline design that would meet process improvement goals.

61 RADIATION PROTECTION AND DOSIMETRY↗