Search NASA⌕ Search

SEARCH · Search NASA

Results for “high level application”

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 271 records · Page 15

Strategic need for a multi-purpose thermal hydraulic loop for support of advanced reactor technologies

This report presents a conceptual design for a new high-temperature multi fluid, multi loop test facility for the INL to support thermal hydraulic, materials, and thermal energy storage research for nuclear and nuclear-hybrid applications. In its initial configuration, the facility will include a high-temperature helium loop, a liquid salt loop, and a hot water/steam loop. The three loops will be thermally coupled through an intermediate heat exchanger (IHX) and a secondary heat exchanger (SHX). Research topics to be addressed with this facility include the characterization and performance evaluation of candidate compact heat exchangers such as printed circuit heat exchangers (PCHEs) at prototypical operating conditions, flow and heat transfer issues related to core thermal hydraulics in advanced helium-cooled and salt-cooled reactors, and evaluation of corrosion behavior of new cladding materials and accident-tolerant fuels for LWRs at prototypical conditions. Based on its relevance to advanced reactor systems, the new facility has been named the Advanced Reactor Technology Integral System Test (ARTIST) facility. Research performed in this facility will advance the state of the art and technology readiness level of high temperature intermediate heat exchangers (IHXs) for nuclear applications while establishing the INL as a center of excellence for the development and certification of this technology. The thermal energy storage capability will support research and demonstration activities related to process heat delivery for a variety of hybrid energy systems and grid stabilization strategies. Experimental results obtained from this research will assist in development of reliable predictive models for thermal hydraulic design and safety codes over the range of expected advanced reactor operating conditions. Proposed/existing IHX heat transfer and friction correlations and criteria will be assessed with information on materials compatibility and instrumentation needs. The experimental database will guide development of appropriate predictive methods and be available for code verification and validation (V&V) related to these systems.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

Exploring the Use of Novel Spatial Accelerators in Scientific Applications

Driven by the need to find alternative accelerators which can viably replace GPUs in next-generation Supercomputing systems, this paper proposes a methodology to enable agile application/hardware co-design. The application-first methodology provides the ability to come up with design of accelerators while working with real-world workloads, available accelerators, and system software. The iterative design process targets a set of kernels in a workload for performance estimates that can prune the design space for later phases of detailed architectural evaluations. To this effect, in this paper, a novel data-parallel device model is introduced that simulates the latency of performance-sensitive operations in an accelerator including data transfers and kernel computation using multi-core CPUs. The use of off-the-shelf simulators, such as pre-RTL simulator Aladdin or multiple tools available for exploring the design of deep neural network accelerators (e.g., Timeloop) is demonstrated for evaluation of various accelerator designs using applications with realistic inputs. Examples of multiple device configurations that are instantiable in a system are explored to evaluate the performance benefit of deploying novel accelerators. The proposed device is integrated with a programming model and system software to potentially explore the impacts of high-level programming languages/compilers and low-level effects such as task scheduling on multiple accelerators. We analyze our methodology for a set of applications that represent high-performance computing (HPC) and graph analytics. The applications include a computational chemistry kernel realized using tensor contractions, triangle counting, GraphSAGE and Breadth-first Search. These applications include kernels such as dense matrix-dense matrix multiplication, sparse matrix-spare matrix multiplication, and sparse matrix-dense vector multiplication. Our results indicate potential performance benefits and insights for system design by including accelerators that realize these kernels along-side general purpose accelerators.

AI, codesign, Accelerated Computing, Modeling and ↗

Software engineering to sustain a high-performance computing scientific application: QMCPACK

We provide an overview of the software engineering efforts and their impact in QMCPACK, a production-level ab-initio Quantum MonteCarlo open-source code targeting high-performance computing (HPC) systems. Aspects included are: (i) strategic expansion ofcontinuous integration (CI) targeting CPU, using GitHub Actions runners, and graphics processing units (GPU) in pre-exascalesystems, using self-hosted hardware; (ii) incremental reduction of memory leaks using sanitizers, (iii) incorporation of Dockercontainers for CI and reproducibility, and (iv) refactoring efforts to improve maintainability, testing coverage, and memory lifetime management. We quantify the value of these improvements by providing metrics to illustrate the shift towards a predictive, rather than reactive, sustainable maintenance approach. Our goal, in documenting the impact of these efforts on QMCPACK, is to contribute to the body of knowledge on the importance of research software engineering (RSE) for the sustainability of community HPC codes and scientific discovery at scale.

Godoy, William↗

Highly efficient terahertz photoconductive metasurface detectors operating at microwatt-level gate powers

Despite their wide use in terahertz (THz) research and technology, the application spectra of photoconductive antenna (PCA) THz detectors are severely limited due to the relatively high optical gating power requirement. This originates from poor conversion efficiency of optical gate beam photons to photocurrent in materials with sub-picosecond carrier lifetimes. Here we show that using an ultra-thin (160 nm), perfectly absorbing low-temperature grown GaAs metasurface as the photoconductive channel drastically improves the efficiency of THz PCA detectors. This is achieved through perfect absorption of the gate beam in a significantly reduced photoconductive volume, enabled by the metasurface. This Letter demonstrates that sensitive THz PCA detection is possible using optical gate powers as low as 5 µW—three orders of magnitude lower than gating powers used for conventional PCA detectors. We show that significantly higher optical gate powers are not necessary for optimal operation, as they do not improve the sensitivity to the THz field. This class of efficient PCA THz detectors opens doors for THz applications with low gate power requirements.

47 OTHER INSTRUMENTATION↗

Comparative Assessment of Experimental Testing of Instrument with an Embedded Digital Device Using Model-Based and Conventional Methods

The use of digital equipment has not been widespread in U.S. nuclear facilities. A major contributing factor to this situation arises from regulatory concerns about common cause failures (CCFs) stemming from the use of identical software in redundant systems, questions as to how to quantify the reliability of software-based products, validity of tools that are used for software verification and validation (V&V), and the effectiveness of software QA procedures. Compounding the concerns are sporadic incidents of digital equipment malfunctions in nuclear power plants, including lockups resulting from software memory leaks, feedwater control malfunctions due to excessive traffic on Ethernet networks, software logic errors, and other problems leading to plant trips. Additionally, the limited amount of demonstrated research and a lack of industry consensus on quantitative software reliability methodologies has hindered the widespread adoption of digital I&C in the nuclear industry. In a recent Regulatory Issue Summary (RIS), the U.S. Nuclear Regulatory Commission (NRC) identified concerns about the impact of greater use of instrumentation and control (I&C) equipment with an embedded digital device (EDD) [1]. Specifically, the NRC staff states that increased use of such devices “may increase a facility’s vulnerability to a CCF” (i.e., common-cause failure). The prevailing NRC guidance regarding software CCF (SCCF) identifies two design attributes that are acceptable for eliminating CCF concerns: (1) diversity or (2) testability (specifically, 100% testability). Either solution can result in high costs and remaining licensing uncertainty (i.e., how much diversity is enough? how to ensure test coverage of every possible sequence of device states?). Consequently, many utilities and reactor designers have limited or avoided more extensive use of digital technology to minimize licensing, scheduling, and financial risk. Without development of cost effective qualification methods to satisfy regulatory requirements and address the potential for CCF vulnerability associated with EDDs, the nuclear power industry may not be able to realize the benefits of digital technology achieved by other industries. In order to address these challenges, the U.S. Department of Energy (DOE) is sponsoring this research to assess the prospect of CCF in digital I&C equipment. The purpose of the current research is to develop an effective approach employing science-based methods to resolve concerns about CCF vulnerability that serve to inhibit deployment of advanced instrumentation (e.g., sensors, actuators, microcontrollers) with EDDs in nuclear power applications. The research objectives address the challenge of establishing high levels of safety and reliability assurance needed for the qualification of EDDs (e.g., microprocessors, programmable logic devices) that are subject to software design faults, complex failure modes, and CCF vulnerability. Specific objectives are: (1) assess the regulatory context for treatment of CCF vulnerability a graded approach to their qualification, (3) develop and extend model-based testing methods to enable effective demonstration of whether devices are subject to CCF, which may arise from vulnerabilities introduced at any stage of the design lifecycle, (4) establish a cost-effective testing framework that incorporates automation and test scenario prioritization, and (5) demonstrate the qualification approach through selection and testing of candidate digital device(s). This report documents an assessment of the results from experimental testing of a representative instrument with an embedded digital device. The experiment serves as a proof-of-concept demonstration of a model-based testing (MBT) framework. The objective of MBT is to ensure that an EDD undergoes suitably comprehensive testing to establish that it is not subject to SCCF. The approach by which MBT accomplishes this objective is through development and application of an effective test suite that provides the appropriate coverage to detect the full range of postulated faults that may arise at the requirements, design, and code (implementation) level of the EDD software. The findings of the experiment involve the application of the MBT framework by Ohio State University (OSU) through testing of the selected demonstration instrument (i.e., the prototype smart pressure sensor developed by Virginia Commonwealth University, VCU). In addition, the experiment findings also involve application of a more conventional black-box testing approach that serves as a baseline against which the MBT methodology can be compared. The primary outcome documented in this report is the comparative analysis of the capabilities demonstrated for each method. The results of the analysis demonstrate clear benefits that arise from the MBT methodology. The comparison indicates greater effectiveness from the more comprehensive evidence available through MBT and enhanced efficiency achieved through automation within the MBT framework. The structure of the report addresses the approach and application of each test method as well as the comparative analysis of the experimental results. Chapter 2 gives a summary overview of the demonstration approach. Chapter 3 covers the baseline testing and Chapter 4 documents the model-based testing. Chapter 5 describes the findings of the comparative analysis and presents the conclusions that can be drawn about the effectiveness and efficiency of the MBT methodology.

47 OTHER INSTRUMENTATION↗

A 4 x 4 array module of position-sensitive virtual Frisch-grid CdZnTe detectors for gamma-ray imaging spectrometers

Arrays of position-sensitive virtual Frisch-grid CdZnTe (CZT) detectors offer an economical approach to making high efficiency and high energy resolution gamma cameras for spectroscopy and imaging of radioactive sources. There are many application areas for such instruments including gamma-rays astronomy, medical and industrial imaging, nonproliferation and nuclear safeguards. We present the test results from a 4x4 array module coupled to the front-end ASIC. The array houses 16 detectors made of 6x6x20 mm 3 CZT crystals. Each crystal is encapsulated inside an ultrathin polyester shell and furnished with four 5-mm-wide charge-sensing pads placed near the anode on each side of the detector. The pad’s signals are used to measure X-Y coordinates while the cathode signals give the interaction depths (Z coordinates). Combined together the signals provide 3D position information of the interaction points which can be used to correct the detector response non-uniformity. This allows developers to use standard grade (unselected) CZT crystals while retaining high spectroscopic performance comparable with that of the H3D pixelated detectors. The array’s design provides flexibility to replace individual detectors and extend their sizes, potentially increasing upwards of 4 cm in height. We started development of position-sensitive VFG detectors several years ago and today this technology has reached a high-maturity level and is ready to use in practical applications.

3D charge-loss correction↗

Advances in Uncertainty and Sensitivity Analysis Methods and Applications in GDSA Framework.

The Spent Fuel and Waste Science and Technology (SFWST) Campaign of the U.S. Department of Energy (DOE) Office of Nuclear Energy (NE), Office of Fuel Cycle Technology (FCT) is conducting research and development (R&D) on geologic disposal of spent nuclear fuel (SNF) and high-level nuclear waste (HLW). Two high priorities for SFWST disposal R&D are design concept development and disposal system modeling. These priorities are directly addressed in the SFWST ''Geologic Disposal Safety Assessment'' (GDSA) control account, which is charged with developing a geologic repository system modeling and analysis capability, and the associated software, ''GDSA Framework'', for evaluating disposal system performance for nuclear waste in geologic media. ''GDSA Framework'' is supported by SFWST Campaign and its predecessor the Used Fuel Disposition (UFD) campaign. This report fulfills the GDSA Uncertainty and Sensitivity Analysis Methods work package (SF-20SN01030403) level 3 milestone — ''Advances in Uncertainty and Sensitivity Analysis Methods and Applications in GDSA Framework'' (M3SF-20SN010304032). It presents high level objectives and strategy for development of uncertainty and sensitivity analysis tools, demonstrates uncertainty quantification (UQ) and sensitivity analysis (SA) tools in GDSA Framework in FY20, and describes additional UQ/SA tools whose future implementation would enhance the UQ/SA capability of ''GDSA Framework''. This work was closely coordinated with the other Sandia National Laboratory GDSA work packages: the GDSA Framework Development work package (SF- 2051\101030404), the GDSA Repository Systems Analysis work package (SF-2051\101030405), and the GDSA PFLOTRAN Development work package (SF-20SN01030406). This report builds on developments reported in previous ''GDSA Framework'' milestones, particularly M2SF- 19SNO1030403.

11 NUCLEAR FUEL CYCLE AND FUEL MATERIALS↗

Application of NSUF Capabilities Towards Understanding the Emulation of High Dose Neutron Irradiations with Ion Beams

Development of new materials for current nuclear power plants (NPPs) or materials for new NPPs requires extensive research, development, and deployment programs. Central to these efforts, particularly for core internals, are studies on the microstructural and mechanical changes due to neutron irradiation. These radiation effects studies are typically completed using materials test reactors followed by expensive and time-consuming post-irradiation examinations (PIE). This sluggish approach has severely limited the nuclear materials field to develop new materials or verify materials for new reactor design use. It is also very difficult to achieve the high levels of damage proposed for future reactors. A logical approach to accelerating the field is the application of ion beams to simulate neutron damage, especially for high dose applications. Ion beams have been used for decades to study radiation effects but not to emulate high dose reactor irradiations. In this regard, new (or renewed) methods for ion beam irradiations, new developments in materials characterization, and greater understanding on the governing mechanisms controlling observed radiation effects have led to a need for further studies to refine the technique. A primary challenge is the ability to determine the ion irradiation conditions that produce microstructures and mechanical properties relevant to in-core neutron irradiation conditions. A large, multi-disciplinary, multi-laboratory program deemed the High Fidelity Ion Beam Simulation of High Dose Neutron Irradiation has been formed to assess and overcome the challenges associated with emulating high dose neutron irradiations using ion beams. Central to this program is the use of dual-beam ion irradiations and corresponding neutron irradiations within the BOR-60 fast reactor. These irradiations have produced hundreds of specimens for microstructural and mechanical characterization. A significant issue regarding the samples is the radioactivity of the neutron irradiated specimens as well as the need to characterize both irradiation types using the same equipment to maintain consistency and repeatability. The use of the Nuclear Science User Facilities (NSUF) has been central towards overcoming the mentioned issues and thereby enabling robust PIE in both the ion and neutron irradiated specimens of interest. In particular, the electron microscopes housed within the Low Activation Materials Development and Analysis (LAMDA) facility at Oak Ridge National Laboratory (ORNL) - an NSUF partner facility - has been instrumental in gathering high quality data for analysis. This paper will briefly highlight the PIE efforts using the NSUF capabilities within LAMDA.

36 MATERIALS SCIENCE↗

Uncertainty and Sensitivity Analysis Methods and Applications in the GDSA Framework (FY2021)

The Spent Fuel and Waste Science and Technology (SFWST) Campaign of the U.S. Department of Energy (DOE) Office of Nuclear Energy (NE), Office of Fuel Cycle Technology (FCT) is conducting research and development (R&D) on geologic disposal of spent nuclear fuel (SNF) and high-level nuclear waste (HLW). Two high priorities for SFWST disposal R&D are design concept development and disposal system modeling. These priorities are directly addressed in the SFWST Geologic Disposal Safety Assessment (GDSA) control account, which is charged with developing a geologic repository system modeling and analysis capability, and the associated software, GDSA Framework, for evaluating disposal system performance for nuclear waste in geologic media. GDSA Framework is supported by SFWST Campaign and its predecessor the Used Fuel Disposition (UFD) campaign. This report fulfills the GDSA Uncertainty and Sensitivity Analysis Methods work package (SF-21SN01030404) level 3 milestone, Uncertainty and Sensitivity Analysis Methods and Applications in GDSA Framework (FY2021) (M3SF-21SN010304042). It presents high level objectives and strategy for development of uncertainty and sensitivity analysis tools, demonstrates uncertainty quantification (UQ) and sensitivity analysis (SA) tools in GDSA Framework in FY21, and describes additional UQ/SA tools whose future implementation would enhance the UQ/SA capability of GDSA Framework. This work was closely coordinated with the other Sandia National Laboratory GDSA work packages: the GDSA Framework Development work package (SF-21SN01030405), the GDSA Repository Systems Analysis work package (SF-21SN01030406), and the GDSA PFLOTRAN Development work package (SF-21SN01030407). This report builds on developments reported in previous GDSA Framework milestones, particularly M3SF 20SN010304032.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Optimization of Deep Borehole Systems for HLW Disposal

This is the final report on a project to update and improve the conceptual design of deep boreholes for high level nuclear waste disposal. The effort was concentrated on application to intact US legacy LWR fuel assemblies, but conducted in a way in which straightforward extension to other waste forms, host rock types and countries was preserved. The reference fuel design version consists of a vertical borehole drilled into granitic bedrock, with the uppermost kilometer serving as a caprock zone containing a diverse and redundant series of plugs. There follows a one to two kilometer waste canister emplacement zone having a hole diameter of approximately 40-50 cm. Individual holes are spaced 200-300 m apart to form a repository field. The choice of verticality and the use of a graphite based mud as filler between the waste canisters and the borehole wall liner was strongly influenced by the expectation that retrievability would continue to be emphasized in US and worldwide repository regulatory criteria. An advanced version was scoped out using zinc alloy cast in place to fill void space inside a disposal canister and its encapsulated fuel assembly. This excludes water and greatly improves both crush resistance and thermal conductivity. However the simpler option of using a sand fill was found adequate and is recommended for near-term use. Thermal-hydraulic modeling of the low permeability and porosity host rock and its small (≤ 1%) saline water content showed that vertical convection induced by the waste’s decay heat should not transport nuclides from the emplacement zone up to the biosphere atop the caprock. First order economic analysis indicated that borehole repositories should be cost-competitive with shallower mined repositories. It is concluded that proceeding with plans to drill a demonstration borehole to confirm expectations, and to carry out priority experiments, such as retention and replenishment of in-hole water is in order.

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

NEAR CRITICAL POINT TESTING AND PERFORMANCE RESULTS OF A SCO2 COMPRESSOR FOR A 10MWE BRAYTON CYCLE

Development and commercial acceptance of sCO2 Brayton cycles for power generation applications are growing rapidly as they offer performance advantages over other cycles. To maximize the cycle performance, the compressors are designed to operate near the critical point of the working fluid. At the critical point the fluid properties change rapidly with variations in inlet conditions. This makes it challenging to both accurately predict the performance and guarantee the operability of the compressor, as the behavior is affected by these slight variations in inlet conditions. A full scale 1st stage main compressor for a 10MWe-Class recompression Brayton cycle was built and tested to validate performance and operability in this unique operating range. The compressor was tested in a laboratory environment with additional instrumentation, beyond what is required by PTC-10, to minimize the uncertainty in the measured performance. Complete constant speed characteristics were collected at multiple supercritical points, operating at constant inlet conditions for each speed line covering a range of compressor inlet densities from 400 to 600 kg/m3. Variations in the compressor stage efficiency and choke margin were observed, and the overall operability and stability of the compressor in response to changes in operating condition were also monitored. The compressor was shown to have excellent performance that closely matched the original design prediction. The performance at various inlet conditions showed minimal change in isentropic head coefficient at the design flow, but did show some variation in efficiency and choke margin across the map. These changes in performance were observed to be minimal, and did not affect the stable operation of the compressor. The results demonstrate that a commercial scale sCO2 compressor can operate near the critical point and achieve the high levels of performance and stability required for power generation applications.

pelton, rob↗

Mixed-Precision S/DGEMM Using the TF32 and TF64 Frameworks on Low-Precision AI Tensor Cores

Using NVIDIA graphics processing units (GPUs) equipped with Tensor Cores has enabled the significant acceleration of general matrix multiplication (GEMM) for applications in machine learning (ML) and artificial intelligence (AI) and in high-performance computing (HPC) generally. The use of such power-efficient, specialized accelerators can provide a performance increase between 8 × and 20 ×, albeit with a loss in precision. However, a high level of precision is required in many large scientific and HPC applications, and computing in single or double precision is still necessary for many of these applications to maintain accuracy. Fortunately, mixed-precision methods can be employed to maintain a higher level of numerical precision while also taking advantage of the performance increases from computing with lower-precision AI cores. With this in mind, we extend the state of the art by using NVIDIA’s new TF32 framework. This new framework not only burdens some constraints of the previous frameworks, such as costly 32 16-bit castings but also provides an equivalent precision and performance by using a much simpler approach. We also propose a new framework called TF64 that attempts double-precision arithmetic with low-precision Tensor Cores. Although this framework does not exist yet, we validated the correctness of this idea and achieved an equivalent of 64-bit precision on 32-bit hardware.

Valero Lara, Pedro↗

Potential of the Julia Programming Language for High Energy Physics Computing

Research in high energy physics (HEP) requires huge amounts of computing and storage, putting strong constraints on the code speed and resource usage. To meet these requirements, a compiled high-performance language is typically used; while for physicists, who focus on the application when developing the code, better research productivity pleads for a high-level programming language. A popular approach consists of combining Python, used for the high-level interface, and C++, used for the computing intensive part of the code. A more convenient and efficient approach would be to use a language that provides both high-level programming and high-performance. The Julia programming language, developed at MIT especially to allow the use of a single language in research activities, has followed this path. In this paper the applicability of using the Julia language for HEP research is explored, covering the different aspects that are important for HEP code development: runtime performance, handling of large projects, interface with legacy code, distributed computing, training, and ease of programming. The study shows that the HEP community would benefit from a large scale adoption of this programming language. The HEP-specific foundation libraries that would need to be consolidated are identified.

97 MATHEMATICS AND COMPUTING↗

Scaling and dimensional analysis of acoustic streaming jets

This paper focuses on acoustic streaming free jets. This is to say that progressive acoustic waves are used to generate a steady flow far from any wall. The derivation of the governing equations under the form of a nonlinear hydrodynamics problem coupled with an acoustic propagation problem is made on the basis of a time scale discrimination approach. This approach is preferred to the usually invoked amplitude perturbations expansion since it is consistent with experimental observations of acoustic streaming flows featuring hydrodynamic nonlinearities and turbulence. Experimental results obtained with a plane transducer in water are also presented together with a review of the former experimental investigations using similar configurations. A comparison of the shape of the acoustic field with the shape of the velocity field shows that diffraction is a key ingredient in the problem though it is rarely accounted for in the literature. A scaling analysis is made and leads to two scaling laws for the typical velocity level in acoustic streaming free jets; these are both observed in our setup and in former studies by other teams. We also perform a dimensional analysis of this problem: a set of seven dimensionless groups is required to describe a typical acoustic experiment. We find that a full similarity is usually not possible between two acoustic streaming experiments featuring different fluids. We then choose to relax the similarity with respect to sound attenuation and to focus on the case of a scaled water experiment representing an acoustic streaming application in liquid metals, in particular, in liquid silicon and in liquid sodium. We show that small acoustic powers can yield relatively high Reynolds numbers and velocity levels; this could be a virtue for heat and mass transfer applications, but a drawback for ultrasonic velocimetry.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

UMap: An application-oriented user level memory mapping library

Exploiting the prominent role of complex memories in exascale node architecture, the UMap page fault handler offers new capabilities to access large memory-mapped data sets directly. UMap provides flexible configuration options to customize page handling to each application, including analysis of massive observational and simulation data sets. The high-performance design features I/O decoupling, dynamic load balancing, and application-level controls. Page faults triggered by application threads and processes accessing data mapped to a UMapp’ed region are handled via the Linux userfaultfd protocol, an asynchronous message-oriented kernel-user communication mechanism that avoids the context switch penalty of traditional signal fault handlers. UMap is fully open source. In this paper, we give an overview of the UMap library architecture, its extensible plugin architecture, and the use/performance of UMap in emerging heterogeneous memory hierarchies such as near-node Non-volatile Memory (NVM) and network attached memories. We highlight new capabilities in two pagefault management plugins, the NetworkStore and SparseStore. We demonstrate the integration between UMap and multiple ECP products including Caliper, Metall, ZFP, Mochi, and Ripples.

97 MATHEMATICS AND COMPUTING↗

Fast and Accurate Simulation of the Cray XMT Multithreaded Supercomputer

Irregular applications, such as data mining and analysis or graph-based computations, show unpredictable memory/network access patterns and control structures. Highly multithreaded architectures with large processor counts, like the Cray MTA-1, MTA-2 and XMT, appear to address their requirements better than commodity clusters. However, the research on highly multithreaded systems is currently limited by the lack of adequate architectural simulation infrastructures due to issues such as size of the machines, memory footprint, simulation speed, accuracy and customization. At the same time, Shared-memory MultiProcessors (SMPs) with multi-core processors have become an attractive platform to simulate large scale machines. In this paper, we introduce a cycle-level simulator of the highly multithreaded Cray XMT supercomputer. The simulator runs unmodified XMT applications. We discuss how we tackled the challenges posed by its development, detailing the techniques introduced to make the simulation as fast as possible while maintaining a high accuracy. By mapping XMT processors (ThreadStorm with 128 hardware threads) to host computing cores, the simulation speed remains constant as the number of simulated processors increases, up to the number of available host cores. The simulator supports zero-overhead switching among different accuracy levels at run-time and includes a network model that takes into account contention. On a modern 48-core SMP host, our infrastructure simulates a large set of irregular applications 500 to 2000 times slower than real time when compared to a 128-processor XMT, while remaining within 10\% of accuracy. Emulation is only from 25 to 200 times slower than real time.

Cray XMT↗

Soil application of high-lignin fermentation byproduct to increase the sustainability of liquid biofuel production from crop residues

When digestates from anaerobic digestion of crop residues are added to soil, a considerable body of information indicates that soil organic carbon (SOC) levels are comparable to those when crop residues are left in the field. This occurs although the amount of digestate added to soil is diminished by digestion and implies that digestion increases the proportion of carbon inputs stabilized as SOC. Here we examine the likelihood and implications of these features being manifested for soil application of high lignin-fermentation byproduct (HLFB) from liquid biofuel production. We show that steady-state SOC levels are much less sensitive to crop residue removal with HLFB return than without it, and provide an example supporting the feasibility of foregoing process energy and coproduct revenue when HLFB is returned to the soil. Informed by this review and analysis, we expect with moderate confidence that long-term SOC levels for soils amended with HLFB from some liquid cellulosic biofuel processes will not be substantially lower than those occurring when crop residues are left in the field. We have high confidence that the economically optimum rate of fertilizer nitrogen (N) application and N 2 O emissions will be lower at most sites for HLFB return to the soil than if crop residues were left in the field. We estimate that the per hectare N demand for processing crop residues to liquid biofuels is about a third of the per hectare demand for crop production, giving rise to an opportunity to use N twice and thereby realize cost savings and environmental benefits. These observations support but do not prove the hypothesis that a ‘win-win’ is possible wherein large amounts of liquid biofuel feedstock can be obtained from crop residues while improving the economics and sustainability of food and feed production. A research agenda aimed at exploring and testing this hypothesis is offered.

09 BIOMASS FUELS↗

Software stewardship and advancement of a high-performance computing scientific application: QMCPACK

Here, we provide an overview of the software engineering efforts and their impact in QMCPACK, a production-level ab-initio Quantum Monte Carlo open-source code targeting high-performance computing (HPC) systems. Aspects included are: (i) strategic expansion of continuous integration (CI) targeting CPUs, using GitHub Actions own runners, and NVIDIA and AMD GPUs used in pre-exascale systems, (ii) incremental reduction of memory leaks using sanitizers, (iii) incorporation of Docker containers for CI and reproducibility, and (iv) refactoring efforts to improve maintainability, testing coverage, and memory lifetime management. We quantify the value of these improvements by providing metrics to illustrate the shift towards a predictive, rather than reactive, maintenance approach. Our goal, in documenting the impact of these efforts on QMCPACK, is to contribute to the body of knowledge on the importance of research software engineering (RSE) for the stewardship and advancement of community HPC codes to enable scientific discovery at scale.

97 MATHEMATICS AND COMPUTING↗