Copy of Methodologies for Advance Warning of Compute Cluster Problems via Statistical Analysis: A Case Study (Conference Presentation).
Abstract not provided.
SEARCH · Search NASA
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.
Abstract not provided.
Abstract not provided.
Abstract not provided.
In CFD simulations of flow mixing in a steam generator (SG) during natural circulation, one is faced with the problem of representing the thousands of SG U-tubes. Typically simplifications are made to render the problem computationally tractable. In particular, one or a number of tubes are lumped in one volume which is treated as a single porous medium. This approach dramatically reduces the computational size of the problem and hence simulation time. In this work, we endeavor to investigate the adequacy of this approach by performing two separate simulations of flow in a mock-up with 262 U-tubes, i.e. one in which the porous media model is used for the tube bundle, and another in which the full geometry is represented. In both simulations, the Reynolds Stress (RMS) model of turbulence is used. We show that in steady state conditions, the porous media treatment yields results which are comparable to those of the full geometry representation (temperature distribution, recirculation ratio, hot plume spread, etc). Hence, the porous media approach can be extended with a good degree of confidence to the full scale SG. (authors)
For several decades, computer scientists have sought guidance on how to evolve architectures, languages, and programming models for optimal performance, efficiency, and productivity. Unfortunately, this guidance is most often taken from the existing software/hardware ecosystem. Architects attempt to provide micro-architectural solutions to improve performance on fixed binaries. Researchers tweak compilers to improve code generation for existing architectures and implementations, and they may invent new programming models for fixed processor and memory architectures and computational algorithms. In today's rapidly evolving world of on-chip parallelism, these isolated and iterative improvements to performance may miss superior solutions in the same way gradient descent optimization techniques may get stuck in local minima. In an initial study, we have developed an alternate approach that, rather than starting with an existing hardware/software solution laced with hidden assumptions, defines the computational problems of interest and invites architects, researchers and programmers to implement novel hardware/ software co-designed solutions. Our work builds on the previous ideas of computational dwarfs, motifs, and parallel patterns by selecting a representative set of essential problems for which we provide: An algorithmic description; scalable problem definition; illustrative reference implementations; verification schemes. For simplicity, we focus initially on the computational problems of interest to the scientific computing community but proclaim the methodology (and perhaps a subset of the problems) as applicable to other communities. We intend to broaden the coverage of this problem space through stronger community involvement. Previous work has established a broad categorization of numerical methods of interest to the scientific computing, in the spirit of the NAS Benchmarks, which pioneered the basic idea of a 'pencil and paper benchmark' in the 1990s. The initial result of the more modern study was the seven dwarfs, which was subsequently extended to 13 motifs. These motifs have already been useful in defining classes of applications for architecture-software studies. However, these broad-brush problem statements often miss the nuance seen in individual kernels. For example, the computational requirements of particle methods vary greatly between the naive (but more accurate) direct calculations and the particle-mesh and particle-tree codes. Thus we commenced our study with an enumeration of problems, but then proceeded by providing not only reference implementations for each problem, but more importantly a mathematical definition that allows one to escape iterative approaches to software/hardware optimization. To ensure long term value, we have augmented each of our reference implementations with both a scalable problem generator and a verification scheme. In a paper we have prepared that documents our efforts, we describe in detail this process of problem definition, scalable input creation, verification, and implementation of reference codes for the scientific computing domain. Table 1 enumerates and describes the level of support we've developed for each kernel. We group these important kernels using the Berkeley dwarfs/motifs taxonomy using a red box in the appropriate column. As kernels become progressively complex, they build upon other, simpler computational methods. We note this dependency via orange boxes. After enumeration of the important numerical problems, we created a domain-appropriate high-level definition of each problem. To ensure future endeavors are not tainted by existing implementations, we specified the problem definition to be independent of both computer architecture and existing programming languages, models, and data types. Then, to provide context as to how such kernels productively map to existing architectures, languages and programming models, we produced reference implementations for most of the kernel (see table). These sample codes should be viewed as 'hints,' designed to show how other designers have mapped a problem's operands and operators to existing hardware and software. Since we wanted such implementations to be illustrative, we tried to ensure they were the most straightforward implementation in the easiest to understand languages using familiar architectures. To that end, most of the linear algebra-oriented computations are written in MATLAB using array indexing to process matrices, rather than one-line library calls to compute the same kernel. This ensures that the kernel's computation is explicit and readable in the implementation and not hidden behind a library. For other problems, such as the Barnes-Hut n-Body solver, the implementations were written in pure C, without any supporting library computations. Along this line, we have also created a scalable problem generator to accompany each computation.
Adaptive Computing is an application-agnostic outer loop framework to strategically deploy simulations and experiments to guide decision making for scale-up analysis. Resources are allocated over successive batches, which makes the allocation adaptive to some objective such as optimization or model training. The framework enables the characterization and management of uncertainties associated with predictive models of complex systems when scale-up questions lead to significant model extrapolation. A key advancement of this framework is its integration of multi-fidelity surrogate modeling, uncertainty management, and automated orchestration of various computing and experimentation resources into a single integrated software package. This enables efficient multi-fidelity modeling across multiple computing resources by incorporating real-world constraints such as relative queue times and throughput on individual machines into the multi-fidelity sampling decision. We discuss applications of this framework to problems in the renewable energy space, including biofuels production, material synthesis, perovskite crystal growth, and building electrical loads.
SUMMARY A higher‐order unsplit multi‐dimensional discretization of the diffuse interface model for two‐material compressible flows proposed by R. Saurel, F. Petitpas and R. A. Berry in 2009 is developed. The proposed higher‐order method is based on the concepts of the Multidimensional Optimal Order Detection (MOOD) method introduced in three recent papers for single‐material flows. The first‐order unsplit multi‐dimensional Finite Volume discretization presented by SPB serves as foundation for the development of the higher‐order unlimited schemes. Specific detection criteria along with a novel decrementing algorithm for the MOOD method are designed in order to deal with the complexity of multi‐material flows. Numerically, we compare errors and computational times on several 1D problems (stringent shock tube and cavitation problems) computed on 2D meshes with the second‐ and fourth‐order MOOD methods using a classical MUSCL method as reference. Several simulations of a 2D shocked R22 bubble in the air are also presented on Cartesian and unstructured meshes with the second‐ and fourth‐order MOOD methods, and qualitative comparisons confirm the conclusions obtained with 1D problems. These numerical results demonstrate the robustness of the MOOD approach and the interest of using more than second‐order methods even for locally singular solutions of complex physics models. Copyright © 2014 John Wiley & Sons, Ltd.
Recently, advances in processor architecture have become the driving force for new programming models in the computing industry, as ever newer multicore processor designs with increasing number of cores are introduced on schedules regimented by marketing demands. As a result, collaborative parallel (rather than simply concurrent) implementations of important applications, programming languages, models, and even algorithms have been forced to adapt to these architectures to exploit the available raw performance. We believe that this optimization regime is flawed. In this paper, we present an alternate approach that, rather than starting with an existing hardware/software solution laced with hidden assumptions, defines the computational problems of interest and invites architects, researchers and programmers to implement novel hardware/software co-designed solutions. Our work builds on the previous ideas of computational dwarfs, motifs, and parallel patterns by selecting a representative set of essential problems for which we provide: An algorithmic description; scalable problem definition; illustrative reference implementations; verification schemes. This testbed will enable comparative research in areas such as parallel programming models, languages, auto-tuning, and hardware/software codesign. For simplicity, we focus initially on the computational problems of interest to the scientific computing community but proclaim the methodology (and perhaps a subset of the problems) as applicable to other communities. We intend to broaden the coverage of this problem space through stronger community involvement.
Recently, advances in processor architecture have become the driving force for new programming models in the computing industry, as ever newer multicore processor designs with increasing number of cores are introduced on schedules regimented by marketing demands. As a result, collaborative parallel (rather than simply concurrent) implementations of important applications, programming languages, models, and even algorithms have been forced to adapt to these architectures to exploit the available raw performance. We believe that this optimization regime is flawed. In this paper, we present an alternate approach that, rather than starting with an existing hardware/software solution laced with hidden assumptions, defines the computational problems of interest and invites architects, researchers and programmers to implement novel hardware/software co-designed solutions. Our work builds on the previous ideas of computational dwarfs, motifs, and parallel patterns by selecting a representative set of essential problems for which we provide: An algorithmic description; scalable problem definition; illustrative reference implementations; verification schemes. This testbed will enable comparative research in areas such as parallel programming models, languages, auto-tuning, and hardware/software codesign. For simplicity, we focus initially on the computational problems of interest to the scientific computing community but proclaim the methodology (and perhaps a subset of the problems) as applicable to other communities. We intend to broaden the coverage of this problem space through stronger community involvement.
Superconducting qubits have emerged as a potentially foundational platform technology for addressing complex computational problems deemed intractable with classical computing. Despite recent advances enabling multiqubit designs that exhibit coherence lifetimes on the order of hundreds of μs, material quality and interfacial structures continue to curb device performance. Two-level system defects in the thin superconducting film and adjacent dielectric regions introduce stochastic noise and dissipate electromagnetic energy at the cryogenic operating temperatures. In this study, we utilize time-of-flight secondary ion mass spectrometry to understand the role specific fabrication procedures play in introducing such dissipation mechanisms in these complex systems. We interrogated Nb thin films and transmon qubit structures fabricated through slight modifications in the processing and vacuum conditions. We find that when the Nb film is sputtered onto the Si substrate, oxide and silicide regions are generated at various interfaces. We also observe that impurity species, such as niobium hydrides and carbides, are incorporated within the niobium layer during the subsequent lithographic patterning steps. The formation of these resistive compounds likely impacts the superconducting properties of the Nb thin film. Additionally, we observe the presence of halogen species distributed throughout the patterned thin films. We conclude by hypothesizing the source of such impurities in these structures in an effort to intelligently fabricate superconducting qubits and extend coherence times moving forward.
The development of multimodality preclinical imaging techniques and the rapid growth of realistic computer simulation tools have promoted the construction and application of computational laboratory animal models in preclinical research. Since the early 1990s, over 120 realistic computational animal models have been reported in the literature and used as surrogates to characterize the anatomy of actual animals for the simulation of preclinical studies involving the use of bioluminescence tomography, fluorescence molecular tomography, positron emission tomography, single-photon emission computed tomography, microcomputed tomography, magnetic resonance imaging, and optical imaging. Other applications include electromagnetic field simulation, ionizing and nonionizing radiation dosimetry, and the development and evaluation of new methodologies for multimodality image coregistration, segmentation, and reconstruction of small animal images. This paper provides a comprehensive review of the history and fundamental technologies used for the development of computational small animal models with a particular focus on their application in preclinical imaging as well as nonionizing and ionizing radiation dosimetry calculations. An overview of the overall process involved in the design of these models, including the fundamental elements used for the construction of different types of computational models, the identification of original anatomical data, the simulation tools used for solving various computational problems, and the applications of computational animal models in preclinical research. The authors also analyze the characteristics of categories of computational models (stylized, voxel-based, and boundary representation) and discuss the technical challenges faced at the present time as well as research needs in the future.
Wind turbine blade design is a highly multidisciplinary process that involves aerodynamics, structures, controls, manufacturing, costs, and other considerations. More efficient blade designs can be found by controlling the airfoil cross-sectional shapes simultaneously with the bulk blade twist and chord distributions. Prior work has focused on incorporating panel-based aerodynamic solvers with a blade design framework to allow for airfoil shape control within the design loop in a tractable manner. Including higher fidelity aerodynamic solvers, such as computational fluid dynamics, makes the design problem computationally intractable. In this work, we couple an invertible neural network trained on high-fidelity airfoil aerodynamic data to a turbine design framework to enable the design of airfoil cross sections within a larger blade design problem. We detail the methodology of this coupled framework and showcase its efficacy by aerostructurally redesigning the IEA 15-MW reference wind turbine blade. The coupled approach reduces the cost of energy by 0.9% compared to a more conventional design approach. This work enables the inclusion of high-fidelity aerodynamic data earlier in the design process, reducing cycle time and increasing certainty in the performance of the optimal design.
Identification of geographic ecoregions has long been of interest to environmental scientists and ecologists for identifying regions of similar ecological and environmental conditions. Such classifications are important for predicting suitable species ranges, for stratification of ecological samples, and to help prioritize habitat preservation and remediation efforts. Hargrove and Hoffman (1999, 2009) have developed geographical spatio-temporal clustering algorithms and codes and have successfully applied them to a variety of environmental science domains, including ecological regionalization; environmental monitoring network design; analysis of satellite-, airborne-, and ground-based remote sensing, and climate model-model and model-measurement intercomparison. With the advances in state-of-the-art satellite remote sensing and climate models, observations and model outputs are available at increasingly high spatial and temporal resolutions. Long time series of these high resolution datasets are extremely large in size and growing. Analysis and knowledge extraction from these large datasets are not just algorithmic and ecological problems, but also pose a complex computational problem. This paper focuses on the development of a massively parallel multivariate geographical spatio-temporal clustering code for analysis of very large datasets using tens of thousands processors on one of the fastest supercomputers in the world.
Abstract Ising spin Hamiltonians are often used to encode a computational problem in their ground states. Quantum Annealing (QA) computing searches for such a state by implementing a slow time-dependent evolution from an easy-to-prepare initial state to a low energy state of a target Ising Hamiltonian of quantum spins, H I . Here, we point to the existence of an analytical solution for such a problem for an arbitrary H I beyond the adiabatic limit for QA. This solution provides insights into the accuracy of nonadiabatic computations. Our QA protocol in the pseudo-adiabatic regime leads to a monotonic power-law suppression of nonadiabatic excitations with time T of QA, without any signature of a transition to a glass phase, which is usually characterized by a logarithmic energy relaxation. This behavior suggests that the energy relaxation can differ in classical and quantum spin glasses strongly, when it is assisted by external time-dependent fields. In specific cases of H I , the solution also shows a considerable quantum speedup in computations.
Optimization problems in finance, physics, and computer science are typically very hard to tackle in classical computing; quantum computing could help speed up computations and provide efficient methods for tackling large problems. Typically, to treat a problem with a quantum computer, the optimal solution is cast as the ground state of a diagonal Hamiltonian. Here, we develop a method, called imaginary-time evolution block encoding (ITE-BE), based on a recent imaginary-time algorithm, which requires no variational parameter optimization, as all parameters can be derived analytically from the target Hamiltonian. We also demonstrate that our method can be successfully combined with other quantum algorithms such as the quantum approximate optimization algorithm (QAOA). For illustration, here we study the MaxCut problem. We find that the QAOA ansatz increases the postselection success of ITE-BE, and shallow QAOA circuits, when boosted with ITE-BE, achieve better performance than deeper QAOA circuits. For the special case of the transverse initial state, we adapt our block-encoding scheme to allow for a deterministic application of the first layer of the circuit.
The rapid growth of computing demands driven by scientific computing, data analytics, and artificial intelligence (AI) advancements has exposed the limitations of traditional digital processing systems. These systems are nearing physical energy barriers, making significant gains in energy efficiency increasingly unattainable. As we advance toward post-exascale computing, disruptive approaches are critical to overcoming these limitations. Among emerging analog solutions, biochemical computing offers a transformative path for achieving orders-of-magnitude improvements in energy efficiency. By leveraging the natural optimization capabilities of chemical reaction networks (CRNs), biochemical systems have the potential to meet high-performance computing needs through natural scalability. However, numerous challenges remain, including theoretical limitations in mapping computational problems to CRNs and practical barriers in implementing biochemical computing devices. In this paper, we present a framework for chemical computation using biochemical systems and introduce key components of our approach for energy-efficient scientific computing. We showcase the feasibility of this framework by solving a system of ordinary differential equations by emulating a chemical reservoir device, demonstrating its potential for addressing modern computing challenges. This work lays a foundational step toward harnessing the computational power of chemistry to design energy-efficient, scalable, high-performance next-generation computing systems.
Determining the prime factors of a given number $N$ is a problem that requires super-polynomial time for conventional digital computers. A polynomial-time algorithm was invented by Shor for quantum computers. In this paper, we present experimental data that demonstrate prime factorization using spin-wave interference but without quantum entanglement. Prime factorization includes three major steps. First, a general-type computer calculates the sequence of numbers m k mod($N$), where $N$ is the number to be factorized, m is a randomly chosen positive integer, and k = 1, 2, 3, 4, 5, 6…. Next, the period of the calculated sequence $r$ is determined by exploiting spin-wave interference. Finally, the general-type computer determines the primes based on the obtained $r$. The experiment for period finding was conducted on a six-terminal Y 3 Fe 2 (FeO 4 ) 3 device. We chose number 15 for testing and determined its primes using a sequence of measurements. The obtained experimental data for a micrometer-sized prototype aimed to demonstrate the benefits of using spin-wave devices to solve complex computational problems. Scalability is one of the major strengths inherent in this type of wave-based device, which may provide a route to nanometer-sized logic circuits. We discuss the physical and technological limitations of this approach, which define the maximum size of N and the computational speed. Although this classical approach cannot compete with the quantum algorithm in terms of efficiency, magnonic holographic devices can potentially be used as complementary logic units aimed at speeding up prime factorization for classical computers.
The deep operator network (DeepONet) has shown remarkable potential in solving partial differential equations (PDEs) by mapping between infinite-dimensional function spaces using labeled datasets. However, in scenarios lacking labeled data, the physics-informed DeepONet (PI-DeepONet) approach, which utilizes the residual loss of the governing PDE to optimize the network parameters, faces significant computational challenges, particularly due to the curse of dimensionality. This limitation has hindered its application to high-dimensional problems, making even standard 3D spatial with 1D temporal problems computationally prohibitive. Additionally, the computational requirement increases exponentially with the discretization density of the domain. Here, to address these challenges and enhance scalability for high-dimensional PDEs, we introduce the Separable physics-informed DeepONet (Sep-PI-DeepONet). This framework employs a factorization technique, utilizing sub-networks for individual one-dimensional coordinates, thereby reducing the number of forward passes and the size of the Jacobian matrix required for gradient computations. By incorporating forward-mode automatic differentiation (AD), we further optimize computational efficiency, achieving linear scaling of computational cost with discretization density and dimensionality, making our approach highly suitable for high-dimensional PDEs. We demonstrate the effectiveness of Sep-PI-DeepONet through three benchmark PDE models: the viscous Burgers’ equation, Biot’s consolidation theory, and a parameterized heat equation. Our framework maintains accuracy comparable to the conventional PI-DeepONet while reducing training time by two orders of magnitude. Notably, for the heat equation solved as a 4D problem, the conventional PI-DeepONet was computationally infeasible (estimated 289.35 h), while the Sep-PI-DeepONet completed training in just 2.5 h. These results underscore the potential of Sep-PI-DeepONet in efficiently solving complex, high-dimensional PDEs, marking a significant advancement in physics-informed machine learning.