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.
Conclusions: Emerging computational hardware may fundamentally change optimization algorithm development; Ising Coprocessors are like ALUs or GPUs for optimization!; Based on initial benchmarking efforts D-Wave hardware is not outperforming state-of-the-art alternatives, but seems competitive
The versatility and wide-ranging applicability of the Ising model, originally introduced to study phase transitions in magnetic materials, have made it a cornerstone in statistical physics and a valuable tool for evaluating the performance of emerging computer hardware. Here, we present a novel implementation of the two-dimensional Ising model on Cerebras Wafer-Scale Engine (WSE) – a revolutionary processor that is opening new frontiers in computing. In our deployment of the checkerboard algorithm, we optimized the Ising model to take advantage of the unique WSE architecture. Specifically, we employed a compressed bit representation storing 16 spins on each int16 word, and efficiently distributed the spins over the processing units enabling seamless weak scaling and limiting communications to only immediate neighboring units. Our implementation can handle up to 754 simulations in parallel, achieving an aggregate of over 61.8 trillion flip attempts per second for Ising models with up to 200 million spins. This represents a gain of up to 148 times over previously reported single-devices with a highly optimized implementation on NVIDIA V100 and up to 88 times in productivity compared to NVIDIA H100. Our findings highlight the significant potential of the WSE in scientific computing, particularly in the field of materials modeling.
Imaging opportunities (viewable conjunctions) of Resident Space Objects (RSOs) by satellites are not continuously discovered. We propose to continuously produce and report viewable conjunctions among objects in orbit. Viewable conjunctions are events in space and time when a satellite may favorably view a Resident Space Object (RSO). Favorability is defined by a set of constraints, e.g., solar illumination, distance between observer and target, orbital location for viewable event. Computing viewable conjunctions requires calculation of orbital propagation while considering constraints based on the state vectors of position, velocity, with covariance for both satellite and RSO. We propose two parallel lanes of effort: acceleration and research. The objective of acceleration is to avoid missed opportunities and reduce latency for satellite maneuver requests through continuous prediction and reporting of viewable conjunctions. The effort will begin by deploying currently available software on dedicated systems and continue with optimizing the code for high performance computing hardware. The research lane aims to expand RSO inspection and modeling capabilities. Among our current research ideas are spectral characterization of RSO materials and planning multiple observations to recover RSO 3D form. Computing resources at Oak Ridge National Laboratory (ORNL) are available for the acceleration work. Laika, Maxar conjunction prediction dashboard software, and Bluesim, Maxar orbital propagation software, are expected to be the first software in the acceleration lane. Laike and Bluesim are to be provided by the sponsor, and output will be made accessible through its dashboard. Deliverables will follow a gated schedule to the sponsor. ORNL will provide progressively more robust viewable conjunction assessments from both modelled and actual ephemerides.
For each operating cycle of the Advanced Test Reactor (ATR) at Idaho National Laboratory, a Core Safety Assurance Package (CSAP) is necessary to demonstrate compliance with the safety basis approved by the United State Department of Energy (DOE). Certain computer codes are used in CSAP development, most of them developed in-house. This work describes replacement of a large set of these codes and updates previous work. Replacement of legacy codes is necessary due to computer hardware failure but also has generally improved user-friendliness.
For each operating cycle of the Advanced Test Reactor (ATR) at Idaho National Laboratory, a Core Safety Assurance Package (CSAP) is necessary to demonstrate compliance with the safety basis approved by the United State Department of Energy (DOE). Certain computer codes are used in CSAP development, most of them developed in-house. This work describes replacement of a large set of these codes and updates previous work. Replacement of legacy codes is necessary due to computer hardware failure but also has generally improved user-friendliness.
Computing plays a significant role in all areas of high energy physics. The Snowmass 2021 CompF4 topical group's scope is facilities R&D, where we consider "facilities" as the computing hardware and software infrastructure inside the data centers plus the networking between data centers, irrespective of who owns them, and what policies are applied for using them. In other words, it includes commercial clouds, federally funded High Performance Computing (HPC) systems for all of science, and systems funded explicitly for a given experimental or theoretical program. This topical group report summarizes the findings and recommendations for the storage, processing, networking and associated software service infrastructures for future high energy physics research, based on the discussions organized through the Snowmass 2021 community study.
The rapid deployment of computing hardware different from the traditional CPU+RAM model in data centers around the world mandates a change in the design of event generators for the Large Hadron Collider, in order to provide economically and ecologically sustainable simulations for the high-luminosity era of the LHC. Parton-level event generation is one of the most computationally demanding parts of the simulation and is therefore a prime target for improvements. We present a production-ready leading-order parton-level event generation framework capable of utilizing most modern hardware and discuss its performance in the standard candle processes of vector boson and top-quark pair production with up to five additional jets.
Next-generation space remote sensing systems may be equipped with imaging arrays that sense data at a rate that outstrips the processing capability of any computing hardware that can operate within a satellite’s power budget. This project developed novel convolutional and recurrent neural networks to detect and estimate point-like events amid clutter, and investigated their efficient and accurate implementation on analog in-memory computing systems that are 10-1000× more energy-efficient than digital processors. This project leveraged two memory devices at different levels of technological maturity: a large-scale analog computing prototype using commercial SONOS charge-trap memory, and electrochemical memory (ECRAM) with intrinsic radiation hardness. We experimentally demonstrated end-to-end analog processing of our neural networks on SONOS and characterized the radiation response of both SONOS and ECRAM. We advanced the state-of-the-art in ECRAM precision and reliability, and developed co-design methods to enable accurate long-term operation of SONOS analog accelerators in space radiation environments.
Driven by potential exponential speedups in business, security, and scientific scenarios, interest in quantum computing is surging. This interest feeds the development of quantum computing hardware, but several challenges arise in optimizing application performance for hardware metrics (e.g., qubit coherence and gate fidelity). In this work, we describe a visual analytics approach for analyzing the performance properties of quantum devices and quantum circuit optimization. Our approach allows users to explore spatial and temporal patterns in quantum device performance data and it computes similarities and variances in key performance metrics. Detailed analysis of the error properties characterizing individual qubits is also supported. We also describe a method for visualizing the optimization of quantum circuits. The resulting visualization tool allows researchers to design more efficient quantum algorithms and applications by increasing the interpretability of quantum computations.
Improving the understanding of how computational fluid dynamics (CFD) direct numerical simulations (DNS) of flows in the High Flux Isotope Reactor (HFIR) perform when run in parallel using the high performance computing (HPC) platform Summit at the Oak Ridge Leadership Computing Facility (OLCF) is of particular importance to boost the computational tools used to support HFIR conversion to low enriched fuel (LEU). Evaluation of scaling performance was driven by the increasing importance of graphics processing unit (GPU) usage in HPC, which is becoming the standard for modern supercomputers such as Summit. The desired results are to obtain a strong positive correlation between the computational resources dedicated to a problem and the relative speed-up of the simulation in comparison to a benchmark. This capability will allow substantially improvement in HFIR flow analytical capabilities, specifically when predicting turbulence properties at high Reynolds numbers. The study leverages previous simulation results performed with code PHASTA (finite element) on HPC platforms Cori (NERSC) and Theta (ALCF) [1] with computing options provided in the computing platform OpenFOAM (finite volume) at OLCF. Transitioning from PHASTA to OpenFOAM will (1) eliminate dependence on third-party software for mesh generation and manipulation, (2) reduce resource needs by employing modern architectures, and (3) build expertise for future modeling of HFIR-specific problems like heat transfer in involute geometry, entrance effects, flow structure in channel corners, and so on—all important issues when defining the available thermal margins in the transition to LEU. CPUs and GPUs differ significantly in their architecture and utilization, as discussed in the literature [2]. The most important differences are in the approach to computations and their memory. A single GPU contains a large quantity of cores, enabling it to perform with a much higher throughput than a CPU, but execution requires a different approach. GPU codes execute instructions using the Single-Instruction Multiple-Thread (SIMT) approach in which a single instruction is used for groups of threads called warps. A warp typically consists of 32 threads which must execute the same set of instructions, although on separate threads. Alternately, a CPU has far fewer cores that are much more flexible in their operation, excelling at quickly performing more complex serial computations. This is why GPUs have greater throughput when properly utilized. The second important difference is seen when comparing their memory spaces. Limited memory allocations and CPU–GPU communications cause a significant bottleneck in GPU-accelerated programs. Further study was required to properly take advantage of GPU resources. A comprehensive analysis of code performance and the model-specific features of turbulence constitutes the core of this work. In this study, a DNS simulation of HFIR channel turbulence was performed with the finite volume CFD code OpenFOAM v2112 and CUDA v11.0 on Red Hat Enterprise Linux v8.2. The OpenFOAM installation had AMGx integrated to enable GPU acceleration and utilizes the PETSc4FOAM library. The computational resources and the problem size were scaled on CPU and CPU + GPU architectures to gain a better understanding of the performance of a DNS problem on modern computing hardware. The study aimed to analyze the scaling of the code exclusively on CPUs and then to examine the scaling of the codes with GPU acceleration enabled. Scaling studies included CPU and GPU acceleration on a mesh of varying resolution to analyze the impact of problem size relative to computational resources. In the course of preparing the GPU configuration on Summit, mainly using the AMGX solvers, difficulties were encountered stemming from constant changes resulting from extensive ongoing development activities and the changing environment. This resulted in the inability to complete the GPU portion of the work. The code was compiled and tested, but production runs to assess acceleration were not performed because the used discretional compute time allocation expired as year-end approached. The Summit HPC platform is scheduled for decommissioning in 2024, making it unattractive for future use with Nvidia-based GPUs. Therefore, the work will be moved onto NERSC machines in FY24. An application was prepared and submitted, and sufficient node-hours were awarded to continue the research in the next calendar year. This report summarizes work performed thus far, which mostly focused on CPU OpenFOAM computing.
High Performance Computing (HPC) compliments theory and informs experimental processes in support of NE programs Big Data Analytics capabilities aid in the discovery of science-relevant relationships in data The C520 staff is made up of 19 Staff – Engineering, Basic Sciences, Computer Science, Statistics, Data Science, Technology Computer Computing Hardware Operations include: Falcon is fully utilized with an average 3X over subscription of waiting jobs over the past year Lemhi recently deployed in Feb 2019 Sawtooth procurement underway with Fall 2019 delivery 509 HPC accounts, lab, industry, university users 1.2 PB disk storage, 1.0 PB scratch storage, 0.8 PB disaster recovery storage Collaborative Computing Center nearing completion with approximately 200 staff -- Office space and 200 racks -- Data Center space for computer equipment
For variational algorithms on the near term quantum computing hardware, it is highly desirable to use very accurate ansatze with low implementation cost. Recent studies have shown that the antisymmetrized geminal power (AGP) wavefunction can be an excellent starting point for ansatze describing systems with strong pairing correlations, as those occurring in superconductors. In this work, we show how AGP can be efficiently implemented on a quantum computer with circuit depth, number of CNOTs, and number of measurements being linear in system size. Using AGP as the initial reference, we propose and implement a unitary correlator on AGP and benchmark it on the ground state of the pairing Hamiltonian. Furthermore, the results show highly accurate ground state energies in all correlation regimes of this model Hamiltonian.
y Reverse-mode algorithmic differentiation (AD) is an established method for obtaining adjoint derivatives of computer simulation applications. In computational fluid dynamics (CFD), adjoint derivatives of a cost function output such as drag or lift with respect to design parameters such as surface coordinates or geometry control points are a key ingredient for shape optimization, uncertainty quantification and flow control. The computational cost of CFD applications and their derivatives makes it essential to use high-performance computing hardware efficiently, including multi- and many-core architectures. Nevertheless, OpenMP is not supported in most AD tools, and previously shown methods achieve poor scalability of the derivative code. We present the AD of an OpenMP-parallelized finite volume compressible flow solver for unstructured meshes. Our approach enables us to reuse the parallelization of the original code in the computation of adjoint derivatives. The method works by identifying code segments that can be differentiated in reverse-mode without changing their memory access pattern. The OpenMP parallelization is integrated into the derivative code during the build process in a way that is robust to modifications of the original code and independent of the OpenMP support of the differentiation tool. We show the scalability of our adjoint CFD solver on test cases ranging from thousands to millions of finite volume mesh cells on CPUs with up to 16 threads as well as on an Intel XeonPhi card with 236 threads. We demonstrate that our approach is more practical to implement for production-sized CFD codes and produces more efficient adjoint derivative code than previously shown AD methods.
A (core-collapse) supernova is the final death call of a massive star during which a gigantic amount of a special kind of elementary particles known as neutrinos are released for a brief period of time. An earlier work by the PI and his collaborators has shown that the dense neutrino gas deep inside the supernova can experience "flavor transformation" or "oscillations" in a collective fashion, a macroscopic quantum phenomenon spanning over distances of 100-1000 kilometers. During this process neutrinos of different energies and propagation directions change their identities simultaneously. This phenomenon may be a key to understanding the mechanism of supernova explosion and the origin of many of the chemical elements found on the earth which were produced in the supernovae exploded in the past. The previous understanding of neutrino oscillations in supernovae was based on a stationary 3-dimensional model with 1 spatial dimension and 2 momentum dimensions. In this "neutrino bulb model" neutrinos are emitted from a spherically symmetric surface and they are decoupled from the motion of matter and the production of chemical elements. In order to accurately model the neutrino oscillations in supernovae and determine its physical consequences, sophisticated 7-dimensional (1 temporal dimension + 3 spatial dimensions + 3 momentum dimensions) modelling of neutrino transport must be performed self-consistently together with the hydrodynamic simulation of the flow of energy and matter inside the supernova and with the calculation of the production and destruction of chemical elements through a large nucleosynthesis network. Although it will be long journey to achieve the lofty goal, our project has made important progress in three different fronts towards this goal. In the first front we explored new computing hardware on which the next generation simulation codes may be run. We ported the neutrino bulb code to the platform of Graphics Processing Unit (GPU) and then later developed a new bulb code that can be run on supercomputers equipped with the Intel Xeon CPUs and/or the Xeon Phi coprocessors. In the second front we developed new algorithms that can be used towards computing neutrino oscillations in multi-dimensional models. We have developed a moment expansion method which is approximately 10 times more efficient than the old algorithm. We have also developed numerical codes to solve neutrino oscillations in toy models with two spatial dimensions. In the last front we used analytic methods to study neutrino oscillations in more sophisticated models, e.g. of multiple spatial dimensions and/or with time dependence. By analyzing the neutrino oscillations in toy models with two spacial dimensions we found that the approximate spacial translation symmetry, which is similar to the spherical symmetry in the neutrino bulb model, can be broken spontaneously as neutrino oscillations begin. By analyzing a time-dependent neutrino bulb model we found that certain modes of neutrino oscillations that are suppressed in the stationary model can become active in the time-dependent model. We have also developed theories which give insights to how ordinary matter may affect neutrino oscillations in supernovae and other astrophysical environments.
Here, a review of current trends in scientific computing reveals a broad shift to open-source and higher-level programming languages such as Python and growing career opportunities over the next decade. Open-source modeling tools accelerate innovation in equation-based and data-driven applications. Significant resources have been deployed to develop data-driven tools (PyTorch, TensorFlow, Scikit-learn) from tech companies that rely on machine learning services to meet business needs while keeping the foundational tools open. Open-source equation-based tools such as Pyomo, CasADi, Gekko, and JuMP are also gaining momentum according to user community and development pace metrics. Integration of data-driven and principles-based tools is emerging. New compute hardware, productivity software, and training resources have the potential to radically accelerate progress. However, long-term support mechanisms are still necessary to sustain the momentum and maintenance of critical foundational packages.
We introduce an efficient distributed sequence parallel approach for training transformer-based deep learning image segmentation models. The neural network models are comprised of a combination of a Vision Transformer encoder with a convolutional decoder to provide image segmentation mappings. The utility of the distributed sequence parallel approach is especially useful in cases where the tokenized embedding representation of image data are too large to fit into standard computing hardware memory. To demonstrate the performance and characteristics of our models trained in sequence parallel fashion compared to standard models, we evaluate our approach using a 3D MRI brain tumor segmentation dataset. We show that training with a sequence parallel approach can match standard sequential model training in terms of convergence. Furthermore, we show that our sequence parallel approach has the capability to support training of models that would not be possible on standard computing resources.
Simulating quantum field theories on a quantum computer is one of the most exciting fundamental physics applications of quantum information science. Dynamical time evolution of quantum fields is a challenge that is beyond the capabilities of classical computing, but it can teach us important lessons about the fundamental fabric of space and time. Whether we may answer scientific questions of interest using near-term quantum computing hardware is an open question that requires a detailed simulation study of quantum noise. Here we present a large scale simulation study powered by a multi-node implementation of qsim using the Google Cloud Platform. We additionally employ newly-developed GPU capabilities in qsim and show how Tensor Processing Units -- Application-specific Integrated Circuits (ASICs) specialized for Machine Learning -- may be used to dramatically speed up the simulation of large quantum circuits. We demonstrate the use of high performance cloud computing for simulating $\mathbb{Z}_2$ quantum field theories on system sizes up to 36 qubits. We find this lattice size is not able to simulate our problem and observable combination with sufficient accuracy, implying more challenging observables of interest for this theory are likely beyond the reach of classical computation using exact circuit simulation.
As quantum chemistry calculations deal with molecular systems of increasing size, the memory requirement to store electron-repulsion integrals (ERIs) greatly outpaces the physical memory available in computing hardware. Here, the Cholesky decomposition of ERIs provides a convenient yet accurate technique to reduce the storage requirement of integrals. Recent developments of a two-step algorithm have drastically reduced the memory operation (MOP) count, leaving the floating operation (FLOP) count as the last frontier of cost reduction in the Cholesky ERI algorithm. In this report, we introduce a dynamic integral tracking, reusing, and compression/elimination protocol embedded in the two-step Cholesky ERI method. Benchmark studies suggest that this technique becomes particularly advantageous when the basis set consists of many computationally expensive high-angular-momentum basis functions. With this dynamic ERI improvement, the Cholesky ERI approach proves to be a highly efficient algorithm with minimal FLOP and MOP count.