Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Accelerators, Computing”

Search indexed NASA NTRS and DOE OSTI research on propulsion, heat transfer, battery materials and energy systems. Follow report and document links to the original sources.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 271 records · Page 15

Hardware-accelerated Components for Hybrid Computing Systems

We present a study on the use of component technology for encapsulating platform-specific hardwareaccelerated algorithms on hybrid HPC systems. Our research shows that component technology can have significant benefits from a software engineering pointof- view to increase encapsulation, portability and reduce or eliminate platform dependence for hardwareaccelerated algorithms. As a demonstration of this concept, we discuss our experience in designing, implementing and integrating an FPGA-accelerated kernel for Polygraph, an application in computational proteomics.

97 MATHEMATICS AND COMPUTING↗

A computational model for exploring particle acceleration during reconnection in macroscale systems

A new computational model suitable for exploring the self-consistent production of energetic electrons during magnetic reconnection in macroscale systems is presented. The equations are based on the recent discovery that parallel electric fields are ineffective drivers of energetic particles during reconnection so that the kinetic scales which control the development of such fields can be ordered out of the equations. The resulting equations consist of a magnetohydrodynamic (MHD) backbone with the energetic component represented by macro-particles described by the guiding center equations. Crucially, the energetic component feeds back on the MHD equations so that the total energy of the MHD fluid and the energetic particles is conserved. The equations correctly describe the firehose instability, whose dynamics plays a key role in throttling reconnection and in controlling the spectra of energetic particles. The results of early tests of the model, including the propagation of Alfvén waves in a system with pressure anisotropy and the growth of firehose modes, establish that the basic algorithm is stable and produces reliable physics results in preparation for further benchmarking with particle-in-cell models of reconnection.

70 PLASMA PHYSICS AND FUSION TECHNOLOGY↗

Andes Data Analysis System at the Oak Ridge Leadership Computing Facility

Andes is a (704)-node commodity-type Linux® cluster. Each of Andes’s 704 nodes contain two 16-core 3.0 GHz AMD EPYC 7302 processors with AMD’s Simultaneous Multithreading (SMT) Technology and 256GB of main memory. Andes also has nine large memory GPU nodes. These nodes each have 1TB of main memory and two NVIDIA K80 GPUs with two 14-core 2.30 GHz Intel Xeon processors with HT Technology.

AMD EPYC↗

Computational Materials Science and Chemistry: Accelerating Discovery and Innovation through Simulation-Based Engineering and Science

This report is based on a SC Workshop on Computational Materials Science and Chemistry for Innovation on July 26-27, 2010, to assess the potential of state-of-the-art computer simulations to accelerate understanding and discovery in materials science and chemistry, with a focus on potential impacts in energy technologies and innovation. The urgent demand for new energy technologies has greatly exceeded the capabilities of today's materials and chemical processes. To convert sunlight to fuel, efficiently store energy, or enable a new generation of energy production and utilization technologies requires the development of new materials and processes of unprecedented functionality and performance. New materials and processes are critical pacing elements for progress in advanced energy systems and virtually all industrial technologies. Over the past two decades, the United States has developed and deployed the world's most powerful collection of tools for the synthesis, processing, characterization, and simulation and modeling of materials and chemical systems at the nanoscale, dimensions of a few atoms to a few hundred atoms across. These tools, which include world-leading x-ray and neutron sources, nanoscale science facilities, and high-performance computers, provide an unprecedented view of the atomic-scale structure and dynamics of materials and the molecular-scale basis of chemical processes. For the first time in history, we are able to synthesize, characterize, and model materials and chemical behavior at the length scale where this behavior is controlled. This ability is transformational for the discovery process and, as a result, confers a significant competitive advantage. Perhaps the most spectacular increase in capability has been demonstrated in high performance computing. Over the past decade, computational power has increased by a factor of a million due to advances in hardware and software. This rate of improvement, which shows no sign of abating, has enabled the development of computer simulations and models of unprecedented fidelity. We are at the threshold of a new era where the integrated synthesis, characterization, and modeling of complex materials and chemical processes will transform our ability to understand and design new materials and chemistries with predictive power. In turn, this predictive capability will transform technological innovation by accelerating the development and deployment of new materials and processes in products and manufacturing. Harnessing the potential of computational science and engineering for the discovery and development of materials and chemical processes is essential to maintaining leadership in these foundational fields that underpin energy technologies and industrial competitiveness. Capitalizing on the opportunities presented by simulation-based engineering and science in materials and chemistry will require an integration of experimental capabilities with theoretical and computational modeling; the development of a robust and sustainable infrastructure to support the development and deployment of advanced computational models; and the assembly of a community of scientists and engineers to implement this integration and infrastructure. This community must extend to industry, where incorporating predictive materials science and chemistry into design tools can accelerate the product development cycle and drive economic competitiveness. The confluence of new theories, new materials synthesis capabilities, and new computer platforms has created an unprecedented opportunity to implement a "materials-by-design" paradigm with wide-ranging benefits in technological innovation and scientific discovery. The Workshop on Computational Materials Science and Chemistry for Innovation was convened in Bethesda, Maryland, on July 26-27, 2010. Sponsored by the Department of Energy (DOE) Offices of Advanced Scientific Computing Research and Basic Energy Sciences, the workshop brought together 160 experts in materials science, chemistry, and computational science representing more than 65 universities, laboratories, and industries, and four agencies. The workshop examined seven foundational challenge areas in materials science and chemistry: materials for extreme conditions, self-assembly, light harvesting, chemical reactions, designer fluids, thin films and interfaces, and electronic structure. Each of these challenge areas is critical to the development of advanced energy systems, and each can be accelerated by the integrated application of predictive capability with theory and experiment. The workshop concluded that emerging capabilities in predictive modeling and simulation have the potential to revolutionize the development of new materials and chemical processes. Coupled with world-leading materials characterization and nanoscale science facilities, this predictive capability provides the foundation for an innovation ecosystem that can accelerate the discovery, development, and deployment of new technologies, including advanced energy systems. Delivering on the promise of this innovation ecosystem requires the following: Integration of synthesis, processing, characterization, theory, and simulation and modeling. Many of the newly established Energy Frontier Research Centers and Energy Hubs are exploiting this integration. Achieving/strengthening predictive capability in foundational challenge areas. Predictive capability in the seven foundational challenge areas described in this report is critical to the development of advanced energy technologies. Developing validated computational approaches that span vast differences in time and length scales. This fundamental computational challenge crosscuts all of the foundational challenge areas. Similarly challenging is coupling of analytical data from multiple instruments and techniques that are required to link these length and time scales. Experimental validation and quantification of uncertainty in simulation and modeling. Uncertainty quantification becomes increasingly challenging as simulations become more complex. Robust and sustainable computational infrastructure, including software and applications. For modeling and simulation, software equals infrastructure. To validate the computational tools, software is critical infrastructure that effectively translates huge arrays of experimental data into useful scientific understanding. An integrated approach for managing this infrastructure is essential. Efficient transfer and incorporation of simulation-based engineering and science in industry. Strategies for bridging the gap between research and industrial applications and for widespread industry adoption of integrated computational materials engineering are needed.

36 MATERIALS SCIENCE↗

OpenMP for Accelerators

OpenMP [13] is the dominant programming model for shared-memory parallelism in C, C++ and Fortran due to its easy-to-use directive-based style, portability and broad support by compiler vendors. Similar characteristics are needed for a programming model for devices such as GPUs and DSPs that are gaining popularity to accelerate compute-intensive application regions. This paper presents extensions to OpenMP that provide that programming model. Our results demonstrate that a high-level programming model can provide accelerated performance comparable to hand-coded implementations in CUDA.

97 MATHEMATICS AND COMPUTING↗

Fault Diagnosis of Hybrid Computing Systems Using Chaotic-Map Method

Computing systems are becoming increasingly complex with nodes consisting of a combination of multi-core central processing units (CPUs), many integrated core (MIC) and graphics processing unit (GPU) accelerators. These computing units and their interconnections are subject to different classes of hardware and software faults, which should be detected to support mitigation measures. We present the chaotic-map method that uses the exponential divergence and wide Fourier properties of the trajectories, combined with memory allocations and assignments to diagnose component-level faults in these hybrid computing systems. We propose lightweight codes that utilize highly parallel chaotic-map computations tailored to isolate faults in arithmetic units, memory elements and interconnects. The diagnosis module on a node utilizes pthreads to place chaotic-map threads on CPU and MIC cores, and CUDA C and OpenCL kernels on GPU blocks. We present experimental diagnosis results on five multi-core CPUs; one MIC; and, seven GPUs with typical diagnosis run-times under a minute.

Rao, Nageswara↗

Hydrodynamic modeling of plasma channel systems for laser plasma accelerators

Structured plasma channels are an essential technology for driving high-gradient, plasma-based acceleration and control of electron and positron beams for advanced concepts accelerators. Laser and gas technologies can permit the generation of long plasma columns known as hydrodynamic, optically-field-ionized (HOFI) channels, which feature low on-axis densities and steep walls. By carefully selecting the background gas and laser properties, one can generate narrow, tunable plasma channels for guiding high intensity laser pulses. Here, we present on the development of simulations of HOFI channels using the FLASH code, a publicly available radiation hydrodynamics code. We explore sensitivities of the channel evolution to laser profile, intensity, and background gas conditions, and identify relevant scalings with laser intensity through a range of practical channel delays.

Accelerators↗

High-performance computing with quantum processing units

The prospects of quantum computing have driven efforts to realize fully functional quantum processing units (QPUs). Recent success in developing proof-of-principle QPUs has prompted the question of how to integrate these emerging processors into modern high-performance computing (HPC) systems. We examine how QPUs can be integrated into current and future HPC system architectures by accounting for func- tional and physical design requirements. We identify two integration pathways that are differentiated by infrastructure constraints on the QPU and the use cases expected for the HPC system. This includes a tight integration that assumes infrastructure bottlenecks can be overcome as well as a loose integration that as- sumes they cannot. We find that the performance of both approaches is likely to depend on the quantum interconnect that serves to entangle multiple QPUs. As a result, we also identify several challenges in assessing QPU performance for HPC, and we consider new metrics that capture the interplay between system architecture and the quantum parallelism underlying computational performance.

97 MATHEMATICS AND COMPUTING↗

SST-GPU: An Execution -Driven CUDA Kernel Scheduler and Streaming-Multiprocessor Compute Model

Programmable accelerators have become commonplace in modern computing systems. Advances in programming models and the availability of massive amounts of data have created a space for massively parallel acceleration where the context for thousands of concurrent threads are resident on-chip. These threads are grouped and interleaved on a cycle-by-cycle basis among several massively parallel computing cores. The design of future supercomputers relies on an ability to model the performance of these massively parallel cores at scale. To address the need for a scalable, decentralized GPU model that can model large GPUs, chiplet-based GPUs and multi-node GPUs, this report details the first steps in integrating the open-source, execution driven GPGPU-Sim into the SST framework. The first stage of this project, creates two elements: a kernel scheduler SST element accepts work from SST CPU models and schedules it to an SM-collection element that performs cycle-by-cycle timing using SSTs Mem Hierarchy to model a flexible memory system.

97 MATHEMATICS AND COMPUTING↗

Accelerating Climate Simulations Through Hybrid Computing

Unconventional multi-core processors (e.g., IBM Cell B/E and NYIDIDA GPU) have emerged as accelerators in climate simulation. However, climate models typically run on parallel computers with conventional processors (e.g., Intel and AMD) using MPI. Connecting accelerators to this architecture efficiently and easily becomes a critical issue. When using MPI for connection, we identified two challenges: (1) identical MPI implementation is required in both systems, and; (2) existing MPI code must be modified to accommodate the accelerators. In response, we have extended and deployed IBM Dynamic Application Virtualization (DAV) in a hybrid computing prototype system (one blade with two Intel quad-core processors, two IBM QS22 Cell blades, connected with Infiniband), allowing for seamlessly offloading compute-intensive functions to remote, heterogeneous accelerators in a scalable, load-balanced manner. Currently, a climate solar radiation model running with multiple MPI processes has been offloaded to multiple Cell blades with approx.10% network overhead.

Zhou, Shujia↗

An algorithm for online optimization of accelerators

We developed a general algorithm for online optimization of accelerator performance, i.e., online tuning, using the performance measure as the objective function. This method, named robust conjugate direction search (RCDS), combines the conjugate direction set approach of Powell's method with a robust line optimizer which considers the random noise in bracketing the minimum and uses parabolic fit of data points that uniformly sample the bracketed zone. Moreover, it is much more robust against noise than traditional algorithms and is therefore suitable for online application. Simulation and experimental studies have been carried out to demonstrate the strength of the new algorithm.

43 PARTICLE ACCELERATORS↗

DRAGON: breaking GPU memory capacity limits with direct NVM access

Heterogeneous computing with accelerators is growing in importance in high performance computing (HPC). Recently, application datasets have expanded beyond the memory capacity of these accelerators, and often beyond the capacity of their hosts. Meanwhile, nonvolatile memory (NVM) storage has emerged as a pervasive component in HPC systems because NVM provides massive amounts of memory capacity at affordable cost. Currently, for accelerator applications to use NVM, they must manually orchestrate data movement across multiple memories and this approach only performs well for applications with simple access behaviors. To address this issue, we developed DRAGON, a solution that enables all classes of GP-GPU applications to transparently compute on terabyte datasets residing in NVM. DRAGON leverages the page-faulting mechanism on the recent NVIDIA GPUs by extending capabilities of CUDA Unified Memory (UM). Our experimental results show that DRAGON transparently expands memory capacity and obtain additional speedups via automated I/O and data transfer overlapping.

Markthub, Pak↗

Accelerating Floating-Point Computations with Intel AMX

Intel AMX is a built-in component of recent Intel CPU architectures, first supported by the Intel Sapphire Rapids in 2023, that enables efficient dense matrix multiplications using mixed precision with low-precision data types. The popularity of mixed-precision algorithms has grown recently, primarily due to their use on GPUs to enhance the efficiency of HPC applications, particularly for the training of large language models. The availability of mixed precision on CPUs represents a cost-effective solution for applications where high speed is not critical. This report shows how to use the Intel AMX accelerator through examples in C++ and Python. The examples will focus on mixed-precision floating-point operations obtained by the use of bfloat16 (or BF16) to accelerate code in single precision. We employ a bottom-up methodology, starting from specific register instructions (TMUL operation) to higher-level applications in libraries such as Intel MKL, PyTorch, and TensorFlow, ensuring a comprehensive understanding of the accelerator's potential. Additionally, we provide insights into the expected performance gains when leveraging the accelerator on the Kestrel HPC machine at the National Renewable Energy Laboratory.

97 MATHEMATICS AND COMPUTING↗

Accelerating Scientific Applications With SambaNova Reconfigurable Dataflow Architecture

Here, our exploratory work finds that the SambaNova Reconfigurable Dataflow Architecture (RDA) along with the SambaFlow software stack provides for an attractive system and solution to accelerate AI for science workloads. We have observed the efficacy of using the system with a diverse set of science applications and reasoned their suitability for performance gains over traditional hardware. As the Data-Scale system provides for a very large memory capacity, the system can be used to train models that typically do not fit in a GPU. The architecture also provides for deeper integration with upcoming supercomputers at the Argonne Leadership Computing Facility (ALCF), a US Department of Energy Office of Science user facility, to help advance science insights.

97 MATHEMATICS AND COMPUTING↗

A Novel Approach for Computing Rigid Body Motion Using Linear Accelerations

Here, a novel approach is presented for computing general rigid body motion based on a few known linear accelerations. This method utilizes linear acceleration data obtained from three distinct points on the body, all within a body-fixed reference frame. The only requirement is that the three chosen points must not be collinear. A system of differential-algebraic equations is derived, combining principles of rigid body kinematics with theory of the rotation group SO(3). These equations provide a framework for numerically computing various motion parameters, including angular velocity, angular acceleration, body orientation, velocity field, acceleration field, and displacement field. By numerically solving this system of equations, we can fully characterize rigid body motion in three-dimensional space. A numerical example is provided to demonstrate the practical implementation and efficacy of the proposed technique, illustrating its potential for accurate motion computation in various applications.

42 ENGINEERING↗

Accelerating Asymptotically Exact MCMC for Computationally Intensive Models via Local Approximations

We construct a new framework for accelerating Markov chain Monte Carlo in posterior sampling problems where standard methods are limited by the computational cost of the likelihood, or of numerical models embedded therein. Our approach introduces local approximations of these models into the Metropolis–Hastings kernel, borrowing ideas from deterministic approximation theory, optimization, and experimental design. Previous efforts at integrating approximate models into inference typically sacrifice either the sampler’s exactness or efficiency; our work seeks to address these limitations by exploiting useful convergence characteristics of local approximations. We prove the ergodicity of our approximate Markov chain, showing that it samples asymptotically from the exact posterior distribution of interest. We describe variations of the algorithm that employ either local polynomial approximations or local Gaussian process regressors. Our theoretical results reinforce the key observation underlying this article: when the likelihood has some local regularity, the number of model evaluations per Markov chain Monte Carlo (MCMC) step can be greatly reduced without biasing the Monte Carlo average. Numerical experiments demonstrate multiple order-of-magnitude reductions in the number of forward model evaluations used in representative ordinary differential equation (ODE) and partial differential equation (PDE) inference problems, with both synthetic and real data. Supplementary materials for this article are available online.

97 MATHEMATICS AND COMPUTING↗

Multi-scale calculation based on dual domain material point method combined with molecular dynamics

This dissertation combines the dual domain material point method (DDMP) with molecular dynamics (MD) in an attempt to create a multi-scale numerical method to simulate materials undergoing large deformations with high strain rates. In these types of problems, the material is often in a thermodynamically non-equilibrium state, and conventional constitutive relations are often not available. In this method, the closure quantities, such as stress, at each material point are calculated from a MD simulation of a group of atoms surrounding the material point. Rather than restricting the multi-scale simulation in a small spatial region, such as phase interfaces, or crack tips, this multi-scale method can be used to consider non-equilibrium thermodynamic e ects in a macroscopic domain. This method takes advantage that the material points only communicate with mesh nodes, not among themselves; therefore MD simulations for material points can be performed independently in parallel. First, using a one-dimensional shock problem as an example, the numerical properties of the original material point method (MPM), the generalized interpolation material point (GIMP) method, the convected particle domain interpolation (CPDI) method, and the DDMP method are investigated. Among these methods, only the DDMP method converges as the number of particles increases, but the large number of particles needed for convergence makes the method very expensive especially in our multi-scale method where we calculate stress in each material point using MD simulation. To improve DDMP, the sub-point method is introduced in this dissertation, which provides high quality numerical solutions with a very small number of particles. The multi-scale method based on DDMP with sub-points is successfully implemented for a one dimensional problem of shock wave propagation in a cerium crystal. The MD simulation to calculate stress in each material point is performed in GPU using CUDA to accelerate the computation. The numerical properties of the multiscale method are investigated as well as the results from this multi-scale calculation are compared of particles needed for convergence makes the method very expensive especially in our multi-scale method where we calculate stress in each material point using MD simulation. To improve DDMP, the sub-point method is introduced in this dissertation, which provides high quality numerical solutions with a very small number of particles. The multi-scale method based on DDMP with sub-points is successfully implemented for a one dimensional problem of shock wave propagation in a cerium crystal. The MD simulation to calculate stress in each material point is performed in GPU using CUDA to accelerate the computation. The numerical properties of the multiscale method are investigated as well as the results from this multi-scale calculation are compared with direct MD simulation results to demonstrate the feasibility of the method. Also, the multi-scale method is applied for a two dimensional problem of jet formation around copper notch under a strong impact.

74 ATOMIC AND MOLECULAR PHYSICS↗