Search NASA⌕ Search

SEARCH · Search NASA

Results for “Writing”

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 757 records · Page 42

Jointly Sponsored Research Program on Energy Related Research

Cooperative Agreements, DE-FC26-08NT43293, DOE-WRI Cooperative Research and Development Program for Fossil Energy-Related Resources began in June 2009. The goal of the Program was to develop, commercialize, and deploy technologies of value to the nation’s fossil and renewable energy industries. To ensure relevancy and early commercialization, the involvement of an industrial partner was encouraged. In that regard, the Program stipulated that a minimum of 20% cost share be achieved in a fiscal year. This allowed WRI to carry a diverse portfolio of technologies and projects at various development technology readiness levels. Depending upon the maturity of the research concept and technology, cost share for a given task ranged from none to as high as 67% (two-thirds). Over the course of the Program, a total of twenty six tasks were proposed for DOE approval. Over the period of performance of the Cooperative agreement, WRI has put in place projects utilizing a total of $\$$7,089,581 in USDOE funds. Against this funding, cosponsors have committed $\$$7,398,476 in private funds to produce a program valued at $\$$14,488,057. Tables 1 and 2 presented at the end of this section is a compilation of the funding for all the tasks conducted under the program. The goal of the Cooperative Research and Development Program for Fossil Energy-Related Resources was to through collaborative research with the industry, develop or assist in the development of innovative technology solutions that will: • Increase the production of United States energy resources – coal, natural gas, oil, and renewable energy resources; • Enhance the competitiveness of United States energy technologies in international markets and assist in technology transfer; • Reduce the nation's dependence on foreign energy supplies and strengthen both the United States and regional economies; and • Minimize environmental impacts of energy production and utilization. Success of the Program can be measured by several criteria. Using the deployment of the federal funding with industrial participation as a performance criterion, over the course of the program, the cosponsors contributed more dollars than the federal funds. As stated earlier, a little more than half of the funding for the Program was derived from industrial partners. The industrial partners also enthusiastically supported the research and development activities with cash contribution of $\$$4,710,372.67, nearly 65% of the required cost share. Work on all of the tasks proposed under the Cooperative Agreement has been completed. This report summarizes and highlights the results from the Program. Under the Cooperative Agreement Program, energy-related tasks emphasized petroleum processing, upgrading and characterization, coal and biomass beneficiation and upgrading, coal combustion systems development including oxy-combustion, emissions monitoring and abatement, coal gasification technologies including gas clean-up and conditioning, hydrogen and liquid fuels production, and the development of technologies for the utilization of renewable energy resources. Environmental-related activities emphasized cleaning contaminated soils using microbial fuel cells, development of processes and sorbents for emissions reduction and recovery of water from power plant flue gas, and biological carbon capture and reuse. Technology enhancement activities included resource characterization studies, development of improved methods, monitors and sensors. In general the goals of the tasks proposed were to enhance competitiveness of U.S. technology, increase production of domestic resources, and reduce environmental impacts associated with energy production and utilization. Technologies being brought to commercialization as a result of the funds provided by the Cooperative Agreement contribute to the overall goals of the USDOE and the nation. Each has broad applicability both within the United States and abroad, thereby helping to enhance the competitiveness of U.S. energy technologies in international markets and assisting in technology transfer. Under the Cooperative Agreement Program, WRI has furthered the development of two different coal upgrading technologies. River Basin Energy technology was scaled-up and demonstrated at a nominal 40 tpd size. Similarly, WRI’s patented mercury removal technology further developed into WRITE Coal technology which was then integrated into oxy-combustion and gasification systems for IGCC and fuels production. Integrated systems with WRITE Coal technology applied at the front end represent substantial environmental and efficiency gains. A variation of the RBE coal upgrading technology is being commercialized as a torrefaction technology for woody biomass. WRI worked with EPRI and NIST to develop and improve mercury calibration standards for emissions monitoring. Working with Chart Energy and Chemicals, WRI scaled-up compact reactor technology for the synthesis of fuels and chemicals from syngas. Compact reactor technology represents a five-fold increase in productivity over conventional reactors making smaller-scale distributed synthesis plants an economical viability. Similarly, WRI's patented mixed alcohol synthesis catalyst production is being scaled-up in collaboration with a commercial catalyst manufacturer.

99 GENERAL AND MISCELLANEOUS↗

MIR Performance Analysis

We provide analysis of Oracle StorageTek T10000 Generation B (T10KB) Media Information Record (MIR) Performance Data gathered over the course of a year from our production High Performance Storage System (HPSS). The analysis shows information in the MIR may be used to improve tape subsystem operations. Most notably, we found the MIR information to be helpful in determining whether the drive or tape was most suspect given a read or write error, and for helping identify which tapes should not be reused given their history of read or write errors. We also explored using the MIR Assisted Search to order file retrieval requests. We found that MIR Assisted Search may be used to reduce the time needed to retrieve collections of files from a tape volume.

97 MATHEMATICS AND COMPUTING↗

The “2T” ion-electron semi-analytic shock solution for code-comparison with xRAGE: A report for FY16

This report documents an effort to generate the semi-analytic "2T" ion-electron shock solution developed in the paper by Masser, Wohlbier, and Lowrie, and the initial attempts to understand how to use this solution as a code-verification tool for one of LANL's ASC codes, xRAGE. Most of the work so far has gone into generating the semi-analytic solution. Considerable effort will go into understanding how to write the xRAGE input deck that both matches the boundary conditions imposed by the solution, and also what physics models must be implemented within the semi-analytic solution itself to match the model assumptions inherit within xRAGE. Therefore, most of this report focuses on deriving the equations for the semi-analytic 1D-planar time-independent "2T" ion-electron shock solution, and is written in a style that is intended to provide clear guidance for anyone writing their own solver.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Improving the representation of soluble iron in climate models

Funding from this grant supported Rachel Sanza, Yan Zhang and partially Samuel Albani. Substantial progress has been made on inclusion of mineralogy, showing the quality of the simulations, and the impact on radiation in the CAM4 and CAM5 (Scanza et al., 2015). In addition, the elemental distribution has been evaluated (and partially supported by this grant) (Zhang et al., 2015), showing that using spatial distributions of mineralogy, improved resperentation of Fe, Ca and Al are possible, compared to the limited available data. A new intermediate complexity soluble iron scheme was implemented in the Bulk Aerosol Model (BAM), which was completed as part of Rachel Scanza’s PhD thesis. Currently Rachel is writing up at least two first author papers describing the general methods and comparison to observations (Scanza et al., in prep.), as well as papers describing the sensitivity to preindustrial conditions and interannual variability. This work lead to the lead PI being asked to write a commentary in Nature (Mahowald, 2013) and two review papers (Mahowald et al., 2014, Mahowald et al., submitted) and contributed to related papers (Albani et al., 2016, Albani et al., 2014, Albani et al., 2015).

54 ENVIRONMENTAL SCIENCES↗

Final Technical Report for CENTRIA's EnergyPeak Building Integrated Photovoltaics ( BIPV ) for Standing Seam Metal Roofing Project (DE-FG36-08GO88131)

Architects have said they will adopt Solar Roofs when they are integrated into roofing systems, building design, plans and specifications, and bidding procedures . Standing Seam Metal Roof manufacturers will answer these architects needs if someone will establish standards, pre-design systems, write specifications, tests systems, meet codes, etc. Energypeak was developed to provide the above so Architects can specify BIPV Standing Seam Roof Systems and Manufacturers can brand, market, sell, manufacture and deliver these systems. EnergyPeak took on the challenge to design BIPV modules using current PV technology, design electrical components, develop wire management systems, test, meet codes, write cad drawings and specifications, commercialize and deliver complete standards to the Industry.

14 SOLAR ENERGY↗

PCF File Format.

PCF files are binary files designed to contain gamma spectra and neutron count rates from radiation sensors. It is the native format for the GAmma Detector Response and Analysis Software (GADRAS) package [1]. It can contain multiple spectra and information about each spectrum such as energy calibration. This document outlines the format of the file that would allow one to write a computer program to parse and write such files.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

UPC++ Programmer’s Guide (v1.0 2017.9)

UPC++ is a C++11 library that provides Asynchronous Partitioned Global Address Space (APGAS) programming. It is designed for writing parallel programs that run efficiently and scale well on distributed-memory parallel computers. The APGAS model is single program, multiple-data (SPMD), with each separate thread of execution (referred to as a rank, a term borrowed from MPI) having access to local memory as it would in C++. However, APGAS also provides access to a global address space, which is allocated in shared segments that are distributed over the ranks. UPC++ provides numerous methods for accessing and using global memory. In UPC++, all operations that access remote memory are explicit, which encourages programmers to be aware of the cost of communication and data movement. Moreover, all remote-memory access operations are by default asynchronous, to enable programmers to write code that scales well even on hundreds of thousands of cores.

97 MATHEMATICS AND COMPUTING↗

Unattended Dual Current Monitor (UDCM) FY17 Summary Report

The UDCM is a low current measurement device designed to record pico-amp to micro-amp currents from radiation detectors. The UDCM is the planned replacement for the IAEA’s obsolete MiniGRAND data acquisition module. Preliminary testing of the UDCM at the IAEA facilities lead to the following recommendations from the IAEA: Increase the measurement range. Lower range by a factor of 5 and upper range by 2 orders of magnitude; Modifications to the web interface; Increase programmable acquisition time to 3600s; Develop a method to handle current offsets and negative current; Error checking when writing data to the uSD card; and Writing BID files along with the currently stored BI0 files.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Validation Data Acquisition in HTTF during PCC Events

A validation experimental campaign was conducted in an Integral Effect Test (IET) facility of High Temperature Gas Reactors (HTGR), the High-Temperature Test Facility (HTTF) at Oregon State University (OSU). The HTTF simulates Depressurized and Pressurized Conduction Cooldowns (DCC and PCC). This campaign required the development of a new laser spectroscopic diagnostic to measure velocity in the challenging conditions of the HTTF: low speed (~1 m/s) gas flows at HTGR prototypical temperature and 1/10th pressure. This was a collaborative effort between co-PIs at The George Washington University (GW), Oregon State University (OSU), and NASA Langley Research Center. The main accomplishments of this project include the record for dynamic range for velocimetry, highest accuracy obtained with this technique, successful deployment in an IET leading to new validation matrix for CFD. These are detailed below and in manuscript appended to this executive summary. For this project, we introduced a new class of laser spectroscopic diagnostics to Thermal-Hydraulics to measure velocity of gases; furthermore, the diagnostic was demonstrated in-situ in an IET during DCC events. In such scenarios, particles used in mainstream techniques, like Particle Image Velocimetry (PIV) are not appropriate as they settle down too rapidly and also contaminate the experimental facility. Molecular tracers stay mixed within the working gas and can seed the flow in a technique called Molecular Tagging Velocimetry (MTV). In MTV a molecular tracer is photo-dissociated by a first (write) laser into radicals or molecules. The pattern created by the write laser is then interrogated with planar laser-induced fluorescence (PLIF), the read pulse(s), which are recorded with a camera. The pattern is probed and matched at two times (interval or probe time, dt), resulting in a time-of-flight velocimetry technique. This project demonstrated a new application range for MTV in gases. MTV has been extensively used for high-speed gas dynamics: molecules do not suffer from inertia associated with particles in regions of high accelerations (such as shock waves). For high-speed flows, the probe time is on the order of 0.1-10 µs. In time-of-flight techniques, the velocity is reconstructed from the recorded displacement, dx, between time interval dt: U = dx/dt. To first order, the resolution of the pattern matching is typically a (fixed) fraction of a pixel; therefore the resolution of the velocity measurement increases with dx and, for a given flow velocity, dt. For the low-speed flows of interest here, this requires long probe time, and therefore finding tracers that are stable for a long time. After an extensive literature survey we identified 6 tracers that would be suitable. We also worked with Spectra Physics to design a custom laser system that is flexible enough to probe these various tracers. The final instrument is made of 4 lasers that are integrated into a compact and movable cart to facilitate the deployment of the instrument off-site. One unique aspect of the laser system is the tunable dye laser (needed to do PLIF of the radicals). By being pumped by two Nd:YAG lasers, this laser is able to provide two pulses with any probe time (dt) for wavelengths between 200 and 900 nm with a linewidth of 3 pm. Accommodating the dual-pulse capability required customizing the pump lasers and the dye laser, making this system unique. Additionally, a custom UV-intensified CCD camera capable of frame-straddling and exposure time as short as 10 ns was acquired. More than two-third of the instrument cost (~$200k) was covered by a cost sharing from the PI’s startup funds. Of the 6 potential tracers identified, two were tested extensively (H2O and N2O) at high-temperature and elevated pressure at GW prior to conducting the tests at OSU. For both tracers, an Ar-F laser (193 nm, 10 ns, 10 mJ/pulse) photo-dissociates trace amounts (0.1 – 5% molar fraction) of the gases into OH and NO. OH is probed at 281.905 nm and NO at 226.186 nm (4-30mJ, 10 ns pulse). NO is very stable in inert environment and we could reconstruct velocities precisely with probe delay times as long at 40 ms, compared to 2 ms for OH, resulting in increased precision. Therefore, for the tests in the HTTF, N2O was selected as the final tracer and its optimal concentration was determined to be 0.5% molar. In the GW laboratory, we obtained a velocity measurement precision of 4 mm/s at conditions mimicking the HTTF tests. We also demonstrated the benefit of molecular tracers over particles and the flexibility of our instrument by continuously measuring velocity from 1000 to 0.1 m/s in a blow down test. This represents a dynamic range on velocity of 80 dB, which is unprecedented for velocimetry! Once the diagnostic was optimized, two two-week campaigns were conducted in the HTTF. This experimental campaign required numerous practical solutions to successfully and rapidly deploy this very sensitive table-top diagnostic in-situ. This was accomplished thanks to very extensive planning from all parties involved in the project. Due to constraints on availability of the HTTF, only adiabatic DCC tests were conducted. Specifically, the velocity was measured at the exit of the hot leg into the Reactor Cavity Simulation Tank (RCST). We were able to record the gas velocity associated with the air ingress for long times (up to one hour). Depending on the initial conditions (density of the gases in the Reactor Pressure Vessel and the RCST) the lock exchange lasted 30-60 s past the valve openings. Large vortices were identified within the shear layer at the interface of the cross flow. These vortices are an indication of shear-driven instability between the counter streams of different density. After the lock-exchange, a small residual flow was still present and its strength decayed exponentially. It finds its source in a buoyancy difference between the gases in both vessels due to the mixing and diffusion of the gases in the cross-over duct and the RPV lower plenum. We did not expect to capture this part of the transient, which would have been less significant (if not negligible) in the presence of a heated core; the precision of our measurement in the HTTF was 6 mm/s, which exceeded our expectations for such a complex test! Using commercial software (Fluent) we simulated the experiment with Computational Fluid Dynamic (CFD) using Reynolds Averaged Navier-Stokes (RANS) turbulence models. We were not able to accurately reproduce the shear layer instability, mixing, and diffusion captured in our experimental data, even with large 3D meshes (>10 M points). We attributed this to the inadequacy of the RANS models for the flow of interest: our experimental Reynolds number was too low and was outside the applicability range of the RANS models. Instead, we should have employed CFD with less modeling, such as Large Eddy (LES) or even Direct Numerical (DNS) simulations. Unfortunately, we did not have the computational resources to do so. Finally, it should be noted that this experimental technique holds great promises for measuring velocity in thermalhydraulics. MTV has an application domain (highlighted in the narrative) that is complimentary to more established (and simpler) techniques, such as PIV. It has attracted the attention of several thermalhydraulics researchers and sponsors to measure velocity in harsh environments. The spectroscopic nature of this technique also enables to expand it to measure temperature and concentration simultaneously to velocity. In lieu of a narrative, 4 detailed journal publications are appended to this executive summary to describe the various development and accomplishments during this project. They include 2 published manuscripts (Experiments in Fluids and Measurement Science and Technology), and two invited contributions following the NURETH-17 conference (Nuclear Science and Technology). Our current Nuclear Engineering and Design draft is not included at this time.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

UPC++ Programmer’s Guide, v1.0-2018.3.0

UPC++ is a C++11 library that provides Partitioned Global Address Space (PGAS) programming. It is designed for writing parallel programs that run efficiently and scale well on distributed-memory parallel computers. The PGAS model is single program, multiple-data (SPMD), with each separate thread of execution (referred to as a rank, a term borrowed from MPI) having access to local memory as it would in C++. However, PGAS also provides access to a global address space, which is allocated in shared segments that are distributed over the ranks. UPC++ provides numerous methods for accessing and using global memory. In UPC++, all operations that access remote memory are explicit, which encourages programmers to be aware of the cost of communication and data movement. Moreover, all remote-memory access operations are by default asynchronous, to enable programmers to write code that scales well even on hundreds of thousands of cores.

97 MATHEMATICS AND COMPUTING↗

UPC++ Programmer’s Guide (v1.0-2019.3.0)

UPC++ is a C++11 library that provides Partitioned Global Address Space (PGAS) programming. It is designed for writing parallel programs that run efficiently and scale well on distributed-memory parallel computers. The PGAS model is single program, multiple-data (SPMD), with each separate constituent process having access to local memory as it would in C++. However, PGAS also provides access to a global address space, which is allocated in shared segments that are distributed over the processes. UPC++ provides numerous methods for accessing and using global memory. In UPC++, all operations that access remote memory are explicit, which encourages programmers to be aware of the cost of communication and data movement. Moreover, all remote-memory access operations are by default asynchronous, to enable programmers to write code that scales well even on hundreds of thousands of cores.

97 MATHEMATICS AND COMPUTING↗

Advanced Data Structures for Improved Cyber Resilience and Awareness in Untrusted Environments: LDRD Report

This report summarizes the work performed under the project "Advanced Data Structures for Improved Cyber Resilience and Awareness in Untrusted Environments." The goal of the project was to design, analyze, and test new data structures for cybersecurity applications. We had two major thrusts: 1) using new/improved write-optimized data structures and/or algorithms to better man- age and analyze high-speed massive-volume cyberstreams, and 2) adding security features to data structures at minimum cost. Write optimization allows data structures to better use secondary memory to store and search a larger amount of useful information. Secondary memory is large compared to main memory, but data movement to and from secondary memory must be carefully managed to run quickly enough to keep up with fast streams. The first thrust included managing cyberstreams in parallel, both multi-threaded and distributed, and improving the benchmarking infrastructure for testing new streaming data structures. We considered both (near) real-time discovery of particular patterns, and improved logging for improved forensics. We considered two kinds of security-feature problem. The first was high-performance history- independent external-memory data structures. These provide certain protections to data if a disk is stolen. We also prove some trade-offs between speed and security in this setting. The second data-security problem is more secure data look-up in secret-shared data bases. This report summarizes the project's major accomplishments, with the background to under- stand these accomplishments. It gathers the abstracts and references for the six refereed publications that have appeared as part of this work. We summarize several accomplishments that will be submitted for publication. We then archive one piece of partial work that is not likely to be published in the near future: validation of history-independent data structure implementations.

97 MATHEMATICS AND COMPUTING↗

UPC++ v1.0 Programmer’s Guide (Revision 2019.9.0)

UPC++ is a C++11 library that provides Partitioned Global Address Space (PGAS) programming. It is designed for writing parallel programs that run efficiently and scale well on distributed-memory parallel computers. The PGAS model is single program, multiple-data (SPMD), with each separate constituent process having access to local memory as it would in C++. However, PGAS also provides access to a global address space, which is allocated in shared segments that are distributed over the processes. UPC++ provides numerous methods for accessing and using global memory. In UPC++, all operations that access remote memory are explicit, which encourages programmers to be aware of the cost of communication and data movement. Moreover, all remote-memory access operations are by default asynchronous, to enable programmers to write code that scales well even on hundreds of thousands of cores.

97 MATHEMATICS AND COMPUTING↗

CogSim: Integrate ML-based Library (LAGER) into Multi-Physics Simulation Code

In LAGER, the Mesh and Simulation History Extract (MASHExtract) package has been used for extracting various types of data from Kull runs. The package itself was written in python and completely enmeshed in the Kull codebase, making it impossible to use with other simulations. Having a version that extends to the Ale3d codebase would necessitate a complete re-write of the package, including a conversion to the C++ language. Extending the MASHExtract package to Ale3d requires a level of simulation independence, much of which would come from handling user requests, data management, and exporting. However, there is still a section of code that would need to be able to handle extracting data from the target codebase. Because of the often insurmountable differences between simulation codebases, the MASHExtract package would need a specific layer to be implemented for each codebase that handles this process of collecting and organizing the requested data. Furthermore, different simulation codebases also have different requirements for various enhancements they rely on—user interface, multi processor distribution, gpu acceleration, etc—all of which would also need to be handled by this layer. In regards to Kull and Ale3d specifically, the Kull implementation (KullMASHExtract) will need to be wrapped in Python, and Ale3d (Ale3dMASHExtract) will need to be wrapped in RAJA. Since Kull is primarily wrapped in SWIG, we should also use this library for our own wrapping to maintain consistency. There are also differences between what metrics are available for extraction. In particular, Kull has a wide array of physics metrics that may not be available in Ale3d, and these differences will need to be accounted for when re-writing MASHExtract.

97 MATHEMATICS AND COMPUTING↗

First-Principles Calculation of Vibrational Mode Lifetimes in Complex Materials (Final Report)

The major goal of this project was to develop and apply the moments-based calculation of frequencies and lifetimes of vibrational modes in solids. Most of the effort went into calculation of frequencies (dependence on temperature) and in comparing to experiment. We were successful in reaching that goal. There were three stage: 1) Test levels of approximation and understand their range of validity; 2a) Write an application program for empirical potentials (Jazz, a wrapper for LAMMPS) and use it to perform calculations for interesting materials; 2b) Write a similar application program for a first-principles electronic structure code; 3) Apply the codes to the study of anharmonicity in a variety of materials (bulk silicon, graphene, fullereness, and carbon nanotubes).

42 ENGINEERING↗

NSUF CoMET Project, Implementation Plan FY2019

The Combined Materials and Experiment Toolkit (CoMET) is an NSUF information management resource designed to give users access to the information and tools necessary for designing meaningful experiments and writing clear, structured proposals. NSUF has a milestone to deploy an experiment design expert system to the NSUF website to support applications for the FY2020 CINR, due on August 30, 2019. To meet this milestone, the CoMET project work for FY2019 focuses on three main objectives: 1.) Integrate the three NSUF web applications by linking data fields in one application with the data found in the others. The NSUF web applications are: The Rapid Turnaround Experiment (RTE) Proposal Solicitation system, the Nuclear Fuels and Materials Library (NFML) and the Nuclear Energy Infrastructure Database (NEID) and the researcher and technical staff account and profile management systems. 2.) Build an Experiment Design Wizard (expert system) that will guide users through the process of writing a statement of work for a CINR or RTE experiment. 3.) Build an Irradiation Effects Calculator to help a PI choose the best neutron irradiation facility for their needs, estimate the time needed to achieve their goals, and estimate the resulting radioactivity of their samples, post-irradiation. These objectives are designed to support the mission of the NSUF and optimize the experiment design portion of the proposal processes. The new researcher database will make finding collaborators easier. The wizard will provide the PI with early NSUF expertise so that their proposed experiments, at the “Letter of Intent” and “Pre-Application” stages are better designed to work within the available NSUF facilities. Utilization of these tools should reduce the proportion of proposals that are judged to be infeasible or “not ready” for NSUF access. It should also improve the efficiency of the NSUF technical leads since the PI will be starting with a better statement of work for the project. The tools will also provide resources for NSUF staff to more efficiently screen and score incoming proposals. This implementation plan will layout the work breakdown structure, activities, resources required, and plan for completion of the activities for CoMET in FY2019.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

UPC++ v1.0 Programmer’s Guide, Revision 2020.10.0

UPC++ is a C++11 library that provides Partitioned Global Address Space (PGAS) programming. It is designed for writing parallel programs that run efficiently and scale well on distributed-memory parallel computers. The PGAS model is single program, multiple-data (SPMD), with each separate constituent process having access to local memory as it would in C++. However, PGAS also provides access to a global address space, which is allocated in shared segments that are distributed over the processes. UPC++ provides numerous methods for accessing and using global memory. In UPC++, all operations that access remote memory are explicit, which encourages programmers to be aware of the cost of communication and data movement. Moreover, all remote-memory access operations are by default asynchronous, to enable programmers to write code that scales well even on hundreds of thousands of cores.

97 MATHEMATICS AND COMPUTING↗