SEARCH · Engineering Papers
Results for “Summit”
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.
Seismic tomography of compressional wave velocity and attenuation structure for Makushin Volcano, Alaska
Located in the central portion of the Alaska-Aleutian subduction zone, Makushin Volcano is among the most active volcanoes in the United States and has been classified as a very high threat by the National Volcano Early Warning System based on its eruptive history and proximity to the City of Unalaska and international air routes. In 2015, five standalone seismic stations and three mini seismic arrays of 15 stations each were deployed on Unalaska Island to supplement the Alaska Volcano Observatory permanent seismic network. This temporary array was operational for one year. Taking advantage of the increased azimuthal coverage and the array's increased earthquake detection capability, we developed body-wave Vp and Vp/Vs seismic images of the velocity structure beneath the volcano. Our results show a complex structure with the upper 4 km of the crust. The shallow high-Vp features possibly delineate remnant magma pathways or conduits. Low-Vp regions are found east of the caldera at approximately 4–8 km depth. This is in agreement with previous seismic tomographic work and InSAR-based geodetic models, which had identified this region as a possible long-term source of magma. A three-dimensional compressional wave attenuation model complemented the velocity models. In general, we observe regions of corresponding low P-wave velocity and high attenuation east of the summit caldera between 5 and 7 km depth, supporting the presence of a melt-rich zone. The absolute Vp values (5.5–6.0 km/s) in this region are consistent with the presence of a magma body with a mafic signature. Here, we suggest that a triggering mechanism that involves mafic injections into a more evolved shallow magma chamber is representative of the current magmatic system below the volcano, similarly to what petrological studies have inferred for past eruptions.
An integrated view of the methane system in the pockmarks at Vestnesa Ridge, 79°N
The Vestnesa Ridge is a NW-SE trending, ~ 100 km-long, 1–2 km-thick contourite sediment section located in the Arctic Ocean, west of Svalbard, at 79°N. Pockmarks align along the ridge summit at water depths of ~ 1200 m; they are ~ 700 m in diameter and ~ 10 m deep relative to the surrounding seafloor. Observations of methane seepage in this area have been reported since 2008. In this report we summarize and integrate the available information to date and report on the first detailed seafloor imaging and camera-guided multicore sampling at two of the most active pockmarks along Vestnesa Ridge, named Lomvi and Lunde. We correlate seafloor images with seismically defined subseafloor structures, providing a geological and ecological context to better understand pockmark formation. Subbottom and seismic surveys, water column imaging, geochemical data and seafloor observations indicate ongoing fluid flow at these pockmarks. Visual inspection and sampling using a high-resolution deep-sea camera and multicorer system show exposed gas hydrate and authigenic carbonate in association with biota within two of these pockmarks. Distributed methane venting at both Lomvi and Lunde supports extensive chemosynthetic communities that include filamentous sulfide-oxidizing bacteria and siboglinid tubeworms, all of which utilize chemical energy provided by the seeping fluids. Focused venting forms shallow gas hydrate, and sustains localized gas discharge from 50-m wide pits within the pockmarks. Cycles of carbonate precipitation and/or exhumation of carbonate deposits are indicated by scattered blocks of various size, pavements, and massive carbonate blocks up to 3 m high. Consistent with other observations along continental margin settings, we show that the extensive authigenic carbonate deposits in the Vestnesa pockmarks represent an important and prolonged methane carbon sink that prevents much of the upwardly flowing methane from reaching the overlying ocean.
Influence of vanadium incorporation on the microstructure, mechanical and tribological properties of Nb–V–Si–N films deposited by reactive magnetron sputtering
Composite Nb–V–Si–N films with various V contents (3.7–13.2 at.%) were deposited by reactive magnetron sputtering and the effects of V content on the microstructure, mechanical and tribological properties of Nb–V–Si–N films were investigated. The results revealed that a three-phase structure, consisting of face-centered cubic (fcc) Nb–V–Si–N, hexagonal close-packed (hcp) Nb–V–Si–N and amorphous Si{sub 3}N{sub 4}, co-exists in the Nb–V–Si–N films and the cubic phase is dominant. The hardness and critical load (L{sub c}) of Nb–V–Si–N films initially increased gradually and reached a summit, then decreased with the increasing V content in the films and the maximum values were 35 GPa and 9.8 N, respectively, at 6.4 at.% V. The combination of V into Nb–Si–N film led to the fracture toughness linearly increasing from 1.11 MPa·m{sup 1/2} at 3.7 at.% V to 1.67 MPa·m{sup 1/2} at 13.2 at.% V. At room temperature (RT), the average friction coefficient decreased from 0.80 at 3.7 at.% V to 0.55 at 13.2 at.% V for the Nb–V–Si–N films. The wear rate of Nb–V–Si–N films initially decreased and then increased after reaching a minimum value of about 6.35 × 10{sup −} {sup 7} mm{sup 3}/N·mm at 6.4 at.% V. As the rise of testing temperature from 200 °C to 600 °C, the average friction coefficient of Nb–V–Si–N films decreased with the increase of the testing temperature regardless of V content. However, the wear rate gradually increased for all films. The average friction coefficient and wear rate at RT and elevated temperatures were mainly influenced by the vanadium oxides with weakly bonded lattice planes. - Highlight: • Fcc-Nb–V–Si–N, hcp-Nb–V–Si–N and amorphous Si{sub 3}N{sub 4} co-existed in the films. • The amount of Si{sub 3}N{sub 4} decreased with increasing V content in the films. • Hardness of Nb–V–Si–N film (6.4 at.%) reached a maximum value of 35 GPa. • Addition of V led to the increase of fracture toughness. • Tribological properties were influenced by lubricant wear debris.
Non-bonding interaction of dual atom catalysts for enhanced oxygen reduction reaction
Here, we demonstrate the design of graphene-supported dual atom catalysts (DACs) for the four-electron oxygen reduction reaction (ORR), by utilizing the non-bonding interaction of counterpart metals (M) that synergistically tune the electronic properties and catalytic activity of the Fe active site in FeMN 6 -DAC and FeMN 8 -DAC systems, where M stands for Fe, Co, Ni, Cu, and Zn. More specifically, for Fe-M distances below 15Å, the non-bonding interaction is significant, making the system act as the DAC. We predicted that FeNiN 6 -DAC and FeNiN 8 -DAC exhibit a low ORR overpotential (η ORR ) of 0.28 V and 0.47 V, respectively, which are at the summits of volcano plots. This low η ORR originates from the high Bader charge transfer coupled with high spin density at the Fe site in both the FeNiN 6 -DAC and FeNiN 8 -DAC systems, which weakens the adsorption of OH* intermediate while enhancing its desorption to H 2 O. Guided by these density functional theory (DFT) computational results, we synthesized FeCoN 8 -DAC and FeNiN 8 -DAC along with N-doped graphene and confirmed their structures with scanning transmission electron microscopy (STEM), X-ray photoelectron spectroscopy (XPS), X-ray absorption near-edge structure (XANES), extended X-ray absorption fine structure (EXAFS), and electron spin resonance (ESR). We verify experimentally the catalytic activities and find that FeNiN 8 -DAC has the low experimental overpotential of 0.39 V with a Tafel slope of 47 mVdec -1 . Based on these results, we propose a DFT-guided strategy to tune the charge transfer and spin population of the active site toward designing DACs for electrochemical ORR.
Development toward a ground-based interferometric phased array for radio detection of high energy neutrinos
Not provided.
Parallel accelerated Custom Correlation Coefficient calculations for genomics applications
The massive quantities of genomic data being made available through gene sequencing techniques are enabling breakthroughs in genomic science in many areas such as medical advances in the diagnosis and treatment of diseases. Analyzing this data, however, is a computational challenge insofar as the computational costs of the relevant algorithms can grow with quadratic, cubic or higher complexity—leading to the need for leadership scale computing. In this paper we describe a new approach to calculations of the Custom Correlation Coefficient (CCC) between Single Nucleotide Polymorphisms (SNPs) across a population, suitable for parallel systems equipped with graphics processing units (GPUs) or Intel Xeon Phi processors. We describe the mapping of the algorithms to accelerated processors, techniques used for eliminating redundant calculations due to symmetries, and strategies for efficient mapping of the calculations to many-node parallel systems. Results are presented demonstrating high per-node performance and near-ideal parallel scalability with rates of more than nine quadrillion (9 × 10 15 ) elementwise comparisons achieved per second with the latest optimized code on the ORNL Titan system, this being orders of magnitude faster than rates achieved using other codes and platforms as reported in the literature. Also it is estimated that as many as 90 quadrillion (90 × 10 15 ) comparisons per second may be achievable on the upcoming ORNL Summit system, an additional 10X performance increase. Finally, in a companion paper we describe corresponding techniques applied to calculations of the Proportional Similarity metric for comparative genomics applications.
A parallel strategy for density functional theory computations on accelerated nodes
Using the Löwdin orthonormalization of tall-skinny matrices as a proxy-app for wavefunction-based Density Functional Theory solvers, we investigate a distributed memory parallel strategy focusing on Graphics Processing Unit (GPU)-accelerated nodes as available on some of the top ranked supercomputers at the present time. Here we present numerical results in the strong limit regime, as it is particularly relevant for First-Principles Molecular Dynamics. We also examine how matrix product-based iterative solvers provide a competitive alternative to dense eigensolvers on GPUs, allowing to push the strong scaling limit of these computations to a larger number of distributed tasks. Our strategy, which relies on replicated Gram matrices and efficient collective communications using the NCCL library, leads to a time-to-solution under 0.5 s for the Löwdin orthonormalization of a tall-skinny matrix of 3000 columns on Summit at Oak Ridge Leadership Facility (OLCF). Given the similarity in computational operations between one iteration of a DFT solver and this proxy-app, this shows the possibility of solving accurately the DFT equations well under a minute for 3000 electronic wave functions, and thus perform First-Principles molecular dynamics of physical systems much larger than traditionally solved on CPU systems.
Immortal rays: Rethinking random ray neutron transport on GPU architectures
The Random Ray Method (TRRM) is a recently developed adaptation of the Method of Characteristics for neutral particle transport simulations. TRRM has demonstrated excellent performance on 3D nuclear reactor benchmark problems using CPU-based compute systems. When porting to GPU-based systems, however, new performance challenges arise that are unique to processors targeting massive fine-grained parallelism. For smaller problems, or for large problems that are domain decomposed across many computational nodes, the problem size per node has insufficient parallelism to saturate GPU node resources, thus greatly limiting speedup. In this study, we report on a newly developed “immortal ray” variant of TRRM. Here, the immortal ray technique exposes significantly more fine-grained parallelism by fundamentally reformulating the numerical details of ray discretization, resulting in performance tradeoffs with significant overall benefit on GPUs. For very small 2D simulation problems we found the new immortal ray variant allowed for up to a 4.4x speedup when run on a single GPU. For larger 3D simulation problems we found the new variant improved strong scaling by 3x when run on the Summit supercomputer.
Porting WarpX to GPU-accelerated platforms
WarpX is a general purpose electromagnetic particle-in-cell code that was originally designed to run on many-core CPU architectures. We describe the strategy, based on the AMReX library, followed to allow WarpX to use the GPU-accelerated nodes on OLCF’s Summit supercomputer, a strategy we believe will extend to the upcoming machines Frontier and Aurora. We summarize the challenges encountered, lessons learned, and give current performance results on a series of relevant benchmark problems.
Towards scaling community detection on distributed-memory heterogeneous systems
Distributed multi-GPU systems pose significant challenges and opportunities for efficient execution of parallel applications. Graph algorithms are generally characterized by irregular memory accesses, low computation to communication ratios, and load balancing problems that are especially hard to address on multi-GPU systems. Graph community detection is an important problem in the emerging domain of graph analytics with numerous applications. In this paper, we present our ongoing work on distributed-memory multi-GPU implementation for graph community detection. Our work parallelizes the widely used (albeit serial) Louvain method on distributed multi-GPU platforms. Supported by an extensive set of experiments on a multi-GPU enabled supercomputer (OLCF Summit) and a single compute node (Nvidia DGX-2®), we demonstrate competitive performance to existing distributed-memory CPU-based implementation, and up to 6.5 better results than Nvidia RAPIDS® CUGRAPH. To the best of our knowledge, this work represents the first effort for community detection on distributed multi-GPU systems. Our approach and related findings can be extended to numerous other iterative graph algorithms on multi-GPU systems.
Low-synch Gram–Schmidt with delayed reorthogonalization for Krylov solvers
The parallel strong-scaling of iterative methods is often determined by the number of global reductions at each iteration. Low-synch Gram-Schmidt algorithms are applied here to the Arnoldi algorithm to reduce the number of global reductions and therefore to improve the parallel strong-scaling of iterative solvers for nonsymmetric matrices such as the GMRES and the Krylov-Schur iterative methods. In the Arnoldi context, the factorization is "left-looking" and processes one column at a time. Among the methods for generating an orthogonal basis for the Arnoldi algorithm, the classical Gram-Schmidt algorithm, with reorthogonalization (CGS2) requires three global reductions per iteration. A new variant of CGS2 that requires only one reduction per iteration is presented and applied to the Arnoldi algorithm. Delayed CGS2 (DCGS2) employs the minimum number of global reductions per iteration (one) for a one-column at-a-time algorithm. The main idea behind the new algorithm is to group global reductions by rearranging the order of operations. DCGS2 must be carefully integrated into an Arnoldi expansion or a GMRES solver. Numerical stability experiments assess robustness for Krylov-Schur eigenvalue computations. Performance experiments on the ORNL Summit supercomputer then establish the superiority of DCGS2 over CGS2.
NekRS, a GPU-accelerated spectral element Navier–Stokes solver
The development of NekRS, a GPU-oriented thermal-fluids simulation code based on the spectral element method (SEM) is described. For performance portability, the code is based on the open concurrent compute abstraction and leverages scalable developments in the SEM code Nek5000 and in libParanumal, which is a library of high-performance kernels for high-order discretizations and PDE-based miniapps. Critical performance sections of the Navier–Stokes time advancement are addressed. Performance results on several platforms are presented here, including scaling to 27,648 V100s on OLCF Summit, for calculations of up to 60B gridpoints.
AI and extreme scale computing to learn and infer the physics of higher order gravitational wave modes of quasi-circular, spinning, non-precessing black hole mergers
We use artificial intelligence (AI) to learn and infer the physics of higher order gravitational wave modes of quasi-circular, spinning, non precessing binary black hole mergers. We trained AI models using 14 million waveforms, produced with the surrogate model NRHybSur3dq8, that include modes up to $\ell$ ≤ 4 and (5,5), except for (4,0) and (4,1), that describe binaries with mass-ratios $\textit{q}$ ≤ 8, individual spins $s^z_{\{1,2\}} \in$[–0.8,0.8], and inclination angle $θ \in$ [0,π]. Our probabilistic AI surrogates can accurately constrain the mass-ratio, individual spins, effective spin, and inclination angle of numerical relativity waveforms that describe such signal manifold. We compared the predictions of our AI models with Gaussian process regression, random forest, k-nearest neighbors, and linear regression, and with traditional Bayesian inference methods through the PyCBC Inference toolkit, finding that AI outperforms all these approaches in terms of accuracy, and are between three to four orders of magnitude faster than traditional Bayesian inference methods. Our AI surrogates were trained within 3.4 hours using distributed training on 1,536 NVIDIA V100 GPUs in the Summit supercomputer.
Programming approaches for scalability, performance, and portability of combustion physics codes
Here, this paper presents the process, strategy, and results associated with porting a typical combustion physics flow solver to current state-of-the-art and future massively-parallel computer architectures. Major focus is placed on the distinct algorithmic structure of these types of codes and how it can be integrated with modern programming paradigms for heterogeneous platforms (i.e., distributed many-core systems with accelerators). An end-to-end case study is presented that exemplifies the process in a generic manner, which then serves as a clear guide with respect to the strategy and best practices leading to a robust and adaptable framework that performs well, is durable over time, is portable, and requires minimal human-effort. This end is accomplished beginning with the use of a mature, validated, structured, multiblock code framework optimized for application of both Large Eddy Simulation (LES) and Direct Numerical Simulation (DNS). This code has been ported to a variety of platforms over the past decade, including most recently the Oak Ridge Leadership Computing Facility’s “Summit” Platform. The experience gained on these multiple platforms provides general insights and thus the results presented are not specific to any one code or platform other than the overarching trend toward distributed many-core systems with accelerators in order to move toward exascale performance. The resultant performance and scalability of the ported code is demonstrated on a real-world application; a state-of-the-art rotating detonation rocket engine simulation that matches the complex geometry and boundary conditions imposed as part of a companion experimental campaign.
Glacial geology of the Hudson Mountains, Amundsen Sea sector, West Antarctica
The Hudson Mountains are situated in the eastern Amundsen Sea sector of the West Antarctic Ice Sheet, adjacent to Pine Island Glacier. They form a volcanic field of 17 stratovolcanoes and parasitic vents, preserved as nunataks. Two former tributaries of Pine Island Glacier (Larter and Lucchitta glaciers) flow through the mountains. Here we present a detailed study of the glacial geology of the area. We describe field observations and measurements of geomorphological features from 15 of the nunataks, meltwater ponds found on the surface of three nunataks and supraglacial features (ice dolines) from two sites near the present grounding line. Together these provide constraints on the past ice sheet extent, flow pathways and thermal regime, and enhance our understanding of the present hydrological regime – all of which are important as context for the observed modern ice sheet behaviour. We find evidence suggesting that all nunataks in the Hudson Mountains were covered by ice during the Last Glacial Maximum (defined here as 26.5-19 ka) and have since deglaciated. Faceted and polished erratic cobbles and boulders of exotic lithologies (syenites, alkali granites, granites, granodiorites, tonalites and gabbros) are numerous and perched on nunatak surfaces. A marked difference between the dominant erratic lithologies on nunataks adjacent to Pine Island Glacier (granite) and Lucchitta Glacier (granodiorite-tonalite) indicates that the ice sheet was transporting clasts from at least two distinct upstream source regions. The similarity in degree of weathering suggests, however, that all the erratics were transported by one phase of (warm-based) glaciation; their presence on or close to the summits of all except one nunatak indicates that the ice sheet during that time was at least 700 m thicker than present. These results are consistent with ice sheet model simulations which suggest that all nunataks in the Hudson Mountains were completely submerged by the Last Glacial Maximum ice sheet.
Comparative evaluation of deep learning workloads for leadership-class systems
Deep learning (DL) workloads and their performance at scale are becoming important factors to consider as we design, develop and deploy next-generation high-performance computing systems. Since DL applications rely heavily on DL frameworks and underlying compute (CPU/GPU) stacks, it is essential to gain a holistic understanding from compute kernels, models, and frameworks of popular DL stacks, and to assess their impact on science-driven, mission-critical applications. At Oak Ridge Leadership Computing Facility (OLCF), we employ a set of micro and macro DL benchmarks established through the Collaboration of Oak Ridge, Argonne, and Livermore (CORAL) to evaluate the AI readiness of our next-generation supercomputers. In this paper, we present our early observations and performance benchmark comparisons between the Nvidia V100 based Summit system with its CUDA stack and an AMD MI100 based testbed system with its ROCm stack. We take a layered perspective on DL benchmarking and point to opportunities for future optimizations in the technologies that we consider.
Process-Based Model Evaluation Using Surface Energy Budget Observations in Central Greenland
Energy exchange at the Greenland ice sheet surface governs surface temperature variability, a factor critical for representing surface melt. Physical processes link driving forces to subsequent surface energy budget responses, including radiative, turbulent, and ground heat fluxes, and ultimately control surface temperature evolution. A reanalysis product (ERA-Interim, ERA-I), operational model (Climate Forecast System version 2, CFSv2), and climate model (Community Earth System Model, CESM) are evaluated using a comprehensive set of surface energy budget observations and process-based relationships obtained at Summit, Greenland. Simulated downwelling longwave radiation is underestimated, which is linked to a deficiency of liquid-bearing clouds. Lower than observed surface albedo, especially in ERA-I, compensates for summer deficiencies in downwelling longwave radiation. In winter, such deficiencies are compensated by an overestimation of the sensible heat flux. Process-based relationships convey that all three models underestimate the response of surface temperature to changes in radiative forcing, primarily due to an overactive ground heat flux response in ERA-I, turbulent heat fluxes in CFSv2, and sensible heat flux in CESM. Cross comparison of three distinct models indicates that the ground heat flux response for ERA-I, CFSv2, and CESM is too high, too low, and comparatively accurate, respectively, signifying the benefit of using an advanced representation of snow properties. Relatively small biases in CESM surface albedo suggest that advances in the representation of cloud microphysics result in more realistic radiative forcing. These results provide insight into model strengths and deficiencies, indicating the importance of representing physical processes when portraying cloud impacts on surface temperature variability.