Search NASA⌕ Search

SEARCH · Search NASA

Results for “hardware performance”

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 1,099 records · Page 61

PDCI Wide-Area Damping Control: PSLF Simulations of the 2017 Test Plan - Light Summer Case

To demonstrate and validate the performance of the wide-area damping control system, the project plans to conduct closed-loop tests on the PDCI in spring/summer 2017. A test plan details the open and closed loop tests to be conducted on the PDCI using the wide-area damping control system. To ensure the appropriate level of preparedness, simulations were performed in order to predict and evaluate any possible unsafe operations before hardware experiments are attempted. This report contains the results from these simulations using the power system dynamics software PSLF (Power System Load Flow, trademark of GE). The simulations use the WECC (Western Electricity Coordinating Council) 2016 light summer and heavy summer base cases and the 2014 dual export base case. Because of the large volume of plots, the results were divided into three reports corresponding to the three base cases. This report contains results from the 2016 light summer base case.

25 ENERGY STORAGE↗

PDCI Wide-Area Damping Control: PSLF Simulations of the 2017 Test Plan - Heavy Summer Case

To demonstrate and validate the performance of the wide-area damping control system, the project plans to conduct closed-loop tests on the PDCI in spring/summer 2017. A test plan details the open and closed loop tests to be conducted on the PDCI using the wide-area damping control system. To ensure the appropriate level of preparedness, simulations were performed in order to predict and evaluate any possible unsafe operations before hardware experiments are attempted. This report contains the results from these simulations using the power system dynamics software PSLF (Power System Load Flow, trademark of GE). The simulations use the WECC (Western Electricity Coordinating Council) 2016 light summer and heavy summer base cases and the 2014 dual export base case. Because of the large volume of plots, the results were divided into three reports corresponding to the three base cases. This report contains results from the 2016 heavy summer base case.

25 ENERGY STORAGE↗

PDCI Wide-Area Damping Control: PSLF Simulations of the 2017 Test Plan - Dual Export Case

To demonstrate and validate the performance of the wide-area damping control system, the project plans to conduct closed-loop tests on the PDCI in spring/summer 2017. A test plan details the open and closed loop tests to be conducted on the PDCI using the wide-area damping control system. To ensure the appropriate level of preparedness, simulations were performed in order to predict and evaluate any possible unsafe operations before hardware experiments are attempted. This report contains the results from these simulations using the power system dynamics software PSLF (Power System Load Flow, trademark of GE). The simulations use the WECC (Western Electricity Coordinating Council) 2016 light summer and heavy summer base cases and the 2014 dual export base case. Because of the large volume of plots, the results were divided into three reports corresponding to the three base cases. This report contains results from the 2014 dual export base case.

25 ENERGY STORAGE↗

Optimizing Dirac Wilson Operator and linear solvers for Intel KNL

Lattice Quantumchromodynamics (QCD) is a powerful tool to numerically access the low energy regime of QCD in a straightforward way with quantifyable uncertainties. In this approach, QCD is discretized on a four dimensional, euclidian space-time grid with millions of degrees of freedom. In modern lattice calculations, most of the work is still spent in solving large, sparse linear systems. This part has two challenges, i.e. optimizing the sparse matrix application as well as BLAS-like kernels used in the linear solver. We are going to present performance optimizations of the Dirac operator (dslash) with and without clover term for recent Intel architectures, including Knights Landing (KNL). Using only a single right hand side, we were able to achieve a good fraction of peak performance for slash, CG and BiCGStab. We will also present a series of experiments we performed on KNL, i.e. running MCDRAM in different modes, enabling or disabling hardware prefetching, using different SoA lengths as well as different sub-NUMA-clustering modes.

Joó, Bálint↗

High Temperature Superconducting Reciprocating Magnetic Separator Final Report

In 2001, under DOE's Superconductivity Partnership Initiative (SPI), E. I. du Pont de Nemours & Co. (Dupont) was awarded a cost-share contract to build a fully functional full-scale model high temperature superconducting reciprocating magnet unit specifically designed for the koalin clay industry. After competitive bidding, American Superconductor (AMSC) was selected to provide the coil for the magnet. Dupont performed the statement of work until September 2004, when it stopped work, with the concurrence of DOE, due to lack of federal funds. DOE had paid all invoices to that point, and Dupont had provided all cost share. At this same time, Dupont determined that this program did not fit with its corporate strategies and notified DOE that it was not interesting in resuming the program when funding became available. AMSC expressed interest in assuming performance of the Agreement to Dupont and DOE, and in March 2005, this project was transferred to AMSC by DOE amendment to the original contract and Novation Agreement between AMSC and Dupont. Design drawings and some hardware components and subassemblies were transferred to AMSC. However, no funding was obligated by DOE and AMSC never performed work on the project. This report contains a summary of the work performed by Dupont up to the September 04 timeframe.

99 GENERAL AND MISCELLANEOUS↗

Using Big Data Technologies for HEP Analysis

The HEP community is approaching an era were the excellent performances of the particle accelerators in delivering collision at high rate will force the experiments to record a large amount of information. The growing size of the datasets could potentially become a limiting factor in the capability to produce scientific results timely and efficiently. Recently, new technologies and new approaches have been developed in industry to answer to the necessity to retrieve information as quickly as possible to analyze PB and EB datasets. Providing the scientists with these modern computing tools will lead to rethinking the principles of data analysis in HEP, making the overall scientific process faster and smoother. In this paper, we are presenting the latest developments and the most recent results on the usage of Apache Spark for HEP analysis. The study aims at evaluating the efficiency of the application of the new tools both quantitatively, by measuring the performances, and qualitatively, focusing on the user experience. The first goal is achieved by developing a data reduction facility: working together with CERN Openlab and Intel, CMS replicates a real physics search using Spark-based technologies, with the ambition of reducing 1 PB of public data in 5 hours, collected by the CMS experiment, to 1 TB of data in a format suitable for physics analysis. The second goal is achieved by implementing multiple physics use-cases in Apache Spark using as input preprocessed datasets derived from official CMS data and simulation. By performing different end-analyses up to the publication plots on different hardware, feasibility, usability and portability are compared to the ones of a traditional ROOT-based workflow.

Cremonesi, Matteo↗

Online and Offline Data Quality Monitoring for the Mu2e Calorimeter

This thesis presents the design, implementation, and validation of a calorimeter Data Quality Monitoring (DQM) toolchain for the Mu2e experiment at Fermilab. Mu2e searches for charged lepton flavor violation via coherent muon-to-electron conversion in the field of an aluminum nucleus, $\mu^- Al \rightarrow e^-Al$, a process whose observation would constitute clear evidence of physics beyond the Standard Model. Achieving target sensitivity requires stringent control of detector performance and data integrity during acquisition, as subtle issues in readout configuration, data formatting, or electronics behavior can compromise reconstruction and bias downstream analyzes. To address these challenges, this work develops a multi-layer DQM approach spanning both raw data validation and reconstructed digi-level diagnostics. At the low level, a fragment analysis component performs word- and bit-field decoding of calorimeter readout blocks, enabling sanity checks of the expected structure and producing detailed error and integrity statistics useful for commissioning and troubleshooting. At the digi level, the CaloDigiDQM analyzer is implemented within the art framework and transforms each CaloDigiCollection into a structured hierarchy of ROOT histograms designed for fast drill-down diagnostics. The module generates coherent monitoring views at global, disk, board, and channel granularity, including occupancy, waveform-derived features (baseline, RMS, peak amplitude and position), and left-right sensor consistency metrics. Detector-aware channel-to-electronics mapping is performed through the conditions system (CaloDAQMap), ensuring that diagnostics remain aligned with hardware identifiers used in operations. For end-to-end testing without reliance on live DAQ data, a synthetic CaloDigi producer is developed to generate realistic waveforms with controlled noise and pulse shapes. The resulting system supports both offline ROOT-file production and online operation, including optional histogram streaming through otsdaq via ots::HistoSender. This toolchain provides a practical and scalable foundation for calorimeter commissioning and stable data collection, enabling early detection of anomalies and reducing operational risk for Mu2e.

Vakulenko, Mark [Drew U.] (ORCID:0009000276197818)↗

Availability Performance and Considerations for LCLS X-Ray FEL at SLAC

The Linac Coherent Light Source (LCLS) is an X-ray Free Electron Laser (FEL) facility located at the SLAC National Accelerator Laboratory. LCLS has been in operation since spring 2009, and it has completed its 3rd user run. LCLS is the first in its class of X-ray FEL user facilities, and presents different availability challenges compared to storage ring light sources. This paper presents recent availability performance of the FEL as well as factors to consider when defining the operational availability figure of merit for user runs. During LCLS [1] user runs, an availability of 95% has been set as a goal. In run III, LCLS photon and electron beam systems achieved availabilities of 94.8% and 96.7%, respectively. The total availability goal can be distributed among subsystems to track performance and identify areas that need attention in order to maintain and improve hardware reliability and operational availability. Careful beam time accounting is needed to understand the distribution of down time. The LCLS complex includes multiple experimental hutches for X-ray science, and each user program has different requirements of a set of parameters that the FEL can be configured to deliver. Since each user may have different criteria for what is considered 'acceptable beam', the quality of the beam must be considered to determine the X-ray beam availability.

43 PARTICLE ACCELERATORS↗

Software enhancements to the IVSEM model of the CTBTO IMS.

Sandia National Laboratories (SNL) developed the Integrated Verification System Evaluation Model (IVSEM) to estimate the performance of the International Monitoring System (IMS) operated by the Comprehensive Nuclear Test Ban Treaty Organization (CTBTO). IVSEM was developed in several phases between 1995 and 2000. The model was developed in FORTRAN with an IDL-based user interface and was compiled for Windows and UNIX operating systems. Continuing interest in this analysis capability, coupled with numerous advances in desktop computer hardware and software since IVSEM was written, enabled significant improvements to IVSEM run-time performance and data analysis capabilities. These improvements were implemented externally without modifying the FORTRAN executables, which had been previously verified. This paper describes the parallelization approach developed to significantly reduce IVSEM run-times and the new test setup and analysis tools developed to facilitate better IVSEM operation.

98 NUCLEAR DISARMAMENT, SAFEGUARDS, AND PHYSICAL P↗

Fast analysis of molecular dynamics trajectories with graphics processing units-Radial distribution function histogramming

The calculation of radial distribution functions (RDFs) from molecular dynamics trajectory data is a common and computationally expensive analysis task. The rate limiting step in the calculation of the RDF is building a histogram of the distance between atom pairs in each trajectory frame. Here we present an implementation of this histogramming scheme for multiple graphics processing units (GPUs). The algorithm features a tiling scheme to maximize the reuse of data at the fastest levels of the GPU's memory hierarchy and dynamic load balancing to allow high performance on heterogeneous configurations of GPUs. Several versions of the RDF algorithm are presented, utilizing the specific hardware features found on different generations of GPUs. We take advantage of larger shared memory and atomic memory operations available on state-of-the-art GPUs to accelerate the code significantly. The use of atomic memory operations allows the fast, limited-capacity on-chip memory to be used much more efficiently, resulting in a fivefold increase in performance compared to the version of the algorithm without atomic operations. The ultimate version of the algorithm running in parallel on four NVIDIA GeForce GTX 480 (Fermi) GPUs was found to be 92 times faster than a multithreaded implementation running on an Intel Xeon 5550 CPU. On this multi-GPU hardware, the RDF between two selections of 1,000,000 atoms each can be calculated in 26.9 s per frame. The multi-GPU RDF algorithms described here are implemented in VMD, a widely used and freely available software package for molecular dynamics visualization and analysis.

97 MATHEMATICS AND COMPUTING↗

cuTS: Scaling Subgraph Isomorphism on Distributed Multi-GPUSystems Using Trie Based Data Structure

Subgraph isomorphism is a pattern-matching algorithm widely used in many domains such as chem-informatics, bioinformatics, databases, and social network analysis. It is computationally expensive and is a proven NP-hard problem. The massive parallelism offered by the GPU hardware is well suited for solving the subgraph isomorphism. However, current GPU implementations are far from the achievable performance. Moreover, the enormous memory requirement of current approaches limits the problem size that can be handled. This work analyzes the fundamental challenges associated with processing the subgraph isomorphism on GPUs and develops an efficient GPU hardware-aware implementation. We also develop a new GPU-friendly trie-based data structure to drastically reduce the intermediate storage space requirement. Hence, our approach runs larger benchmarks than the competitors. We also develop the first distributed sub-graph isomorphism algorithm for GPUs. Our experimental evaluation section demonstrates the efficacy of our approach by comparing the execution time and number of cases that we can handle against the state-of-the-art GPU implementations.

Xiang, Lizhi↗

The Development of a Smart Distribution Grid Testbed for Integrated Information Management Systems

This paper presents a smart distribution grid testbed to test or compare designs of integrated information management systems (I2MSs). An I2MS extracts and synthesizes information from a wide range of data sources to detect abnormal system behaviors, identify possible causes, assess the system status, and provide grid operators with response suggestions. The objective of the testbed is to provide a modeling environment with sufficient data sources for the I2MS design. The testbed includes five information layers and a physical layer; it generates multi-layer chronological data based on actual measurement playbacks or simulated data sets produced by the physical layer. The testbed models random hardware failures, human errors, extreme weather events, and deliberate tampering attempts to allow users to evaluate the performance of different I2MS designs. Initial results of I2MS performance tests showed that the testbed created a close-to-real-world environment that allowed key performance metrics of the I2MS to be evaluated.

99 GENERAL AND MISCELLANEOUS↗

Caustic-Side Solvent Extraction Full-Scale Test

A Full-Scale Test (FST) program was performed by Parsons and its team members General Atomics and Energy Solutions to assess the performance of full-scale centrifugal contactors specified for the Department of Energy Salt Waste Processing Facility (SWPF). The SWPF, to be located at the Savannah River Site (SRS) in Aiken, South Carolina, will remove highly radioactive waste constituents, principally actinides, strontium (Sr), and cesium (Cs) radionuclides, from salt waste solutions currently stored in SRS high-level waste tanks. Caustic-side Solvent Extraction (CSSX) removes Cs from waste feed that has been treated upstream to remove actinides and Sr. CSSX uses a custom solvent to extract Cs from the salt solution in a series of single stage centrifugal contactors. The test system comprised (a) eleven 25.4 cm (10'') full-scale contactors (versus 36 in SWPF) for the extraction, scrub, strip, and wash stages; (b) two solvent recovery coalescers; and (c) the associated hardware and control system, packaged in four skid mounted modules. This paper describes the results of tests performed to define both hydraulic performance parameters (maximum hydraulic capacity and phase carryover) and solvent extraction performance parameters (Cs mass transfer efficiencies) using simulated SWPF waste and actual CSSX solvent. The test results confirmed key design features of the CSSX process and, as a consequence, the use of CSSX in the SWPF. In conclusion: Total throughput was initially limited to 85% of maximum flow during FST. Minor system modifications performed prior to mass transfer testing series resulted in the realization of 100% throughput. The 100% flow equates to slightly more than 35.6 x 10{sup 6} L/yr (9.4 Mgal/yr) of waste processed in SWPF which is anticipated to be the peak plant throughput. To achieve the best hydraulic performance in extraction, it is recommended that the extraction contactors be operated at the highest reasonable speed possible (>2100 rpm). Vibration, hardware limitations, bearing life, and other factors should be considered prior to final selection of extraction contactor speeds in SWPF. In strip (also scrub and wash) aqueous carryover decreased and organic carryover increased as the rotor speeds increased. It is recommended that the strip, scrub, and wash contactors be operated at intermediate speeds (between 1500 and 2100 rpm) to achieve a performance compromise between aqueous and organic carryover. Curved-vane bottom plates showed a significant hydraulic performance (aqueous and organic carryover) advantage over straight-vane bottom plates in extraction. There was no significant mass transfer performance advantage for either plate type in extraction. Thus, curved-vane bottom plates in extraction may be the better option for use in SWPF. There was no significant hydraulic performance difference between the plate types in strip. Straight-vanes provided significantly better mass transfer performance in strip compared to curved-vanes. Based solely on mass transfer performance, straight-vane bottom plates in the strip, scrub, and wash contactors are recommended for use in SWPF. Utilizing straight-vanes in the stripping section, the overall SWPF CSSX performance is expected to meet or exceed the target DF of 40,000 with minimum extraction D{sub Cs} of 10. (authors)

12 MANAGEMENT OF RADIOACTIVE AND NON-RADIOACTIVE W↗

Ultracoatings: Enabling Energy and Power Solutions in High Contact Stress Environments through next-generation Nanocoatings Final Technical Report

A review of current commercially available, industrial-grade, low friction coatings will show that interfacial contact pressures nearing 1GPa (~150ksi) inherently limit surface engineering solutions like WC, TiN, TiAlN, and so forth. Extremely hard coatings, then, are often pursued as the principle path, although they too are not without significant limitations. A majority of these compounds are inherently brittle in nature or may not pair well with their mating substrate. In either case, their durability in high contact stress environments is compromised. In parallel to thin film coatings, many conventional surface treatments do not yield an interface hard enough to withstand extreme stresses under load. New research into advanced, nanocomposite materials like (Ti, Zr)B 2 shows great promise. Bulk compacts of this compound have demonstrated an order of magnitude better wear resistance than current offerings, notably materials like tungsten carbide. At a laboratory level, the (Ti,Zr)B 2 nanocomposite material exhibited abrasive and erosive wear resistance nearly ten times better than existing mixed-phase boride systems. In ASTM abrasion and erosion testing, these new compositions exhibit wear resistance superior to other known advanced materials such as RocTec 500 and 'Borazon' cubic boron nitride. Many significant challenges exist for mass production of (Ti, Zr)B 2 , one of which is the necessary processing technology that is capable of minimizing deleterious impurity phases. Secondly, this material's performance is derived from a synergistic effect of the two materials existing as a single phase structure. While the individual constituents of TiB 2 and ZrB 2 do yield improvements to wear resistance, their singular effects are not as significant. Lastly, deposition of this material on a commercial level requires thorough knowledge of nanocomposite boride solids; the benefits associated with these innovative new materials are just being realized. Advancing this technology, called Ultracoatings, through initial development, scale up, and commercialization to a variety of markets would represent a transformative leap to surface engineering. Several application spaces were considered for immediate implementation of the Ultracoatings technology, including, but not limited to, a drive shaft for an aerospace fuel pump, engine timing components, and dry solids pump hardware for an innovative coal gasifier. The primary focus of the program was to evaluate and screen the performance of the selected (Ti, Zr)B 2 Ultracoatings composition for future development. This process included synthesis of the material for physical vapor deposition, sputtering trials and coating characterization, friction and wear testing on sample coupons, and functional hardware testing. The main project deliverables used to gage the project's adherence to its original objective were: Development of a coating/substrate pairing that exhibits wear rate of 0.1 mg/hour or lower at a 1GPa contact pressure, while achieving a maximum coating cost of $0.10/cm 2 . Demonstrate the aforementioned wear rate in both lubricated and starved lubrication conditions. Although the (Ti, Zr) B 2 coating was not tailored for low friction performance, friction and wear evaluations of the material demonstrated a coefficient of sliding friction as low as 0.09. This suggests that varying the percentage of TiB 2 present in the composite could enhance the materials performance in water-based lubricants. In the aerospace drive shaft application, functional hardware coated with (Ti, Zr)B 2 survived a variety of abuse and long-range durability tests, with contact pressures exceeding 2 GPa. For engine timing components, further work is planned to evaluate the Ultracoatings technology in direct injection and diesel engine conditions. In the final identified application space the dry solids pump hardware, discussions continue on the application of the Ultracoatings technology for those specific components. Full implementation of the technology into the targeted markets equates to a U.S.-based energy savings potential of over 100 trillion BTU per year by 2030. This exceeds the original projection of 60 TBTU/year by 2030.

(Ti,Zr) B2↗

Visualization, understanding, and mitigation of process-induced-membrane irregularities in gas diffusion electrode-based polymer electrolyte membrane fuel cells

Polymer electrolyte membrane fuel cells (PEMFC) show substantial promise for their application in electric vehicles. For large-scale manufacturing of PEMFCs, roll-to-roll coated gas-diffusion-electrodes (GDE) offer certain advantages over other production pathways. Procedures including hot pressing and coating an ionomer overlayer may be necessary for this manufacturing pathway to enable a suitable catalyst layer/membrane interface. The same procedures may potentially introduce membrane irregularities, especially when thin membranes are used. Limited understanding exists regarding if and to what extent such irregularities impact PEMFC performance and lifetime, and therefore be considered defects. In this study, NREL's customized fuel cell hardware that enables quasi in-situ infrared (IR) thermography studies was utilized to visualize spatial hydrogen crossover and identify membrane irregularities that originated from the GDE-based MEA fabrication process. The structure of these membrane irregularities was investigated by scanning electron microscopy (SEM) and its impact on initial H2/air performance was determined. Accelerated stress testing (AST) revealed that these irregularities develop into failure point locations. These results were validated across many MEAs with identified process-induced membrane irregularities. By selecting specific gas diffusion media properties and by fine tuning the MEA hot pressing parameters, the formation of such membrane irregularities was mitigated.

08 HYDROGEN↗

Throughput Analytics of Cloud Networks

A network of virtual machines at cloud server sites connected over virtual IO connections is a flexible, easily deployable, and cost-effective alternative to a physical network infrastructure with dedicated servers connected over leased fiber lines. We study the throughput performance of such a cloud network by collecting measurements over Google Cloud infrastructure spanning multiple continents. To study its ideal performance and impact of packet losses, we utilize its emulation using dedicated servers and connection hardware emulation devices. We compare the measurements over the cloud network to those over its emulation on a testbed. We examine the throughput profiles of both networks as a function of the round trip time and their utilization-concavity coefficients, estimated using measurements for common TCP versions. The throughput profile's concave-convex shape and its coefficient are critical indicators of the network performance, qualitatively and quantitatively, respectively. The results indicate their overall agreement between the production cloud network and its emulation using dedicated connections, and a near optimal throughput performance of the former except for a few under-performing connections. Also, the number of parallel flows is found to be a dominant factor in optimizing the throughput across various conditions and TCP versions.

Phanekham, Derek↗

Performance optimization, modeling and analysis of sparse matrix-matrix products on multi-core and many-core processors

Sparse matrix-matrix multiplication (SpGEMM) is a computational primitive that is widely used in areas ranging from traditional numerical applications to recent big data analysis and machine learning. Although many SpGEMM algorithms have been proposed, hardware specific optimizations for multi- and many-core processors are lacking and a detailed analysis of their performance under various use cases and matrices is not available. In this work, we firstly identify and mitigate multiple bottlenecks with memory management and thread scheduling on Intel Xeon Phi (Knights Landing or KNL). Specifically targeting many-core processors, we develop a hash-table-based algorithm and optimize a heap-based shared-memory SpGEMM algorithm. We examine their performance together with other publicly available codes. Different from the literature, our evaluation also includes use cases that are representative of real graph algorithms, such as multi-source breadth-first search or triangle counting. Our hash-table and heap-based algorithms are showing significant speedups from libraries in the majority of the cases while different algorithms dominate the other scenarios with different matrix size, sparsity, compression factor and operation type. We wrap up in-depth evaluation results and make a recipe to give the best SpGEMM algorithm for target scenario. We build the performance model for hash-table and heap-based algorithms, which supports the recipe. A critical finding is that hash-table-based SpGEMM gets a significant performance boost if the nonzeros are not required to be sorted within each row of the output matrix. Finally, we integrate our implementations into a large-scale protein clustering code named HipMCL, accelerating its SpGEMM kernel by up to 10X and achieving an overall performance boost for the whole HipMCL application by 2.6X.

97 MATHEMATICS AND COMPUTING↗

Evaluating Unified Memory Performance in HIP

Heterogeneous unified memory management between a CPU and a GPU is a major challenge in GPU computing. Recently, unified memory (UM) has been supported by software and hardware components on AMD computing platforms. The support could simplify the complexities of memory management. In this paper, we attempt to have a better understanding of UM by evaluating the performance of UM programs on an AMD MI100 GPU. More specifically, we evaluate data migration using UM against other data transfer techniques for the overall performance of an application, assess the impacts of three commonly used optimization techniques on the kernel execution time of a vector add sample, and compare the performance and productivity of selected benchmarks with and without UM. The performance overhead associated with UM is not trivial, but it can improve programming productivity by reducing lines of code for scientific applications. We aim to present early results and feedback on the UM performance to the vendor.

Jin, Zheming↗