Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Computer architecture”

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

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

At least 1,333 records · Page 74

Using the ATLAS experiment software on heterogeneous resources

With the large dataset expected from 2030 onwards by the HL-LHC at CERN, the ATLAS experiment is reaching the limits of the current data processing model in terms of traditional CPU resources based on x86_64 architectures and an extensive program for software upgrades towards the HL-LHC has been set up. The ARM CPU architecture is becoming a competitive and energy efficient alternative. Accelerators like GPUs are available in any recent HPC. In the past years ATLAS has successfully ported its full data processing and simulation software framework Athena to ARM and has invested significant effort in porting parts of the reconstruction and simulation algorithms to GPUs. We report on the successful usage of the ATLAS experiment offline and online software framework Athena on ARM and GPUs through the PanDA workflow management system at various WLCG sites. Furthermore we report on performance optimizations of the builds for ARM CPUs and the GPU integration efforts. We will discuss performance comparisons of different ARM and x86_64 architectures on WLCG resources and Cloud compute providers like GCP and AWS using ATLAS productions workflows as used in the Hep-Score23 benchmark suite.

73 NUCLEAR PHYSICS AND RADIATION PHYSICS↗

Probing many-body localization on a noisy quantum computer

A disordered quantum system of interacting particles exhibits localized behavior when the disorder is large compared to the interaction strength. Studying this phenomenon on a quantum computer with no, or limited, error correction is challenging because even weak coupling to a thermal environment destroys most signatures of localization. Fortunately, spectral functions of local operators are known to contain features that can survive the presence of noise. In these spectra, discrete peaks and a soft gap at low frequencies compared to the thermal phase indicate localization. Here, we present the computation of spectral functions on a trapped-ion quantum computer for a one-dimensional Heisenberg model with disorder. Further, we design an error-mitigation technique which is effective at removing the noise from the measurement allowing clear signatures of localization to emerge as the disorder increases. Furthermore, we show that spectral functions can serve as a robust and scalable diagnostic of many-body localization on current and future generations of quantum computers.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Simulating non-native cubic interactions on noisy quantum machines

As a milestone for general-purpose computing machines, we demonstrate that quantum processors can be programed to efficiently simulate dynamics that are not native to the hardware. Moreover, on noisy devices without error correction, we show that simulation results are significantly improved when the quantum program is compiled using modular gates instead of a restricted set of standard gates. We demonstrate the general methodology by solving a cubic interaction problem, which appears in nonlinear optics, gauge theories, as well as plasma and fluid dynamics. To encode the non-native Hamiltonian evolution, we decompose the Hilbert space into a direct sum of invariant subspaces in which the nonlinear problem is mapped to a finite-dimensional Hamiltonian simulation problem. Furthermore, in a three-states example, the resultant unitary evolution is realized by a product of approximately 20 standard gates, using which approximately ten simulation steps can be carried out on state-of-the-art quantum hardware before results are corrupted by decoherence. In comparison, the simulation depth is improved by more than an order of magnitude when the unitary evolution is realized as a single cubic gate, which is compiled directly using optimal control. Alternatively, parametric gates may also be compiled by interpolating control pulses. Modular gates thus obtained provide high-fidelity building blocks for quantum Hamiltonian simulations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Impacts of noise and structure on quantum information encoded in a quantum memory

As larger, higher-quality quantum devices are built and demonstrated in quantum information applications, such as quantum computation and quantum communication, the need for high-quality quantum memories to store quantum states becomes ever more pressing. Future quantum devices likely will use a variety of physical hardware, some being used primarily for processing of quantum information and others for storage. Here we study the correlation of the structure of quantum information with physical noise models of various possible quantum memory implementations. Through numerical simulation of different noise models and approximate analytical formulas applied to a variety of interesting quantum states, we provide comparisons between quantum hardware with different structure, including both qubit- and qudit-based quantum memories. Our findings point to simple, experimentally relevant formulas for the relative lifetimes of quantum information in different quantum memories and have relevance to the design of hybrid quantum devices.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Suppressing Quantum Circuit Errors Due to System Variability

We present a quantum circuit optimization technique that takes into account the variability in error rates that is inherent across present-day noisy quantum computing platforms. This method can be run after qubit routing or postcompilation and consists of computing isomorphic subgraphs to input circuits and scoring each using heuristic cost functions derived from system calibration data. Using an independent standard algorithmic test suite, we show that it is possible to recover on average nearly 40% of missing fidelity using better qubit selection via efficient to compute cost functions. We demonstrate additional performance gains by considering qubit placement over multiple quantum processors. The overhead from these tools is minimal with respect to other compilation steps, such as qubit routing, as the number of qubits increases. As such, our method can be used to find qubit mappings for problems at the scale of quantum advantage and beyond.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Pixel-Resolved Long-Context Learning for Turbulence at Exascale: Resolving Small-scale Eddies Toward the Viscous Limit

Turbulence plays a crucial role in multiphysics applications, including aerodynamics, fusion, and combustion. Accurately capturing turbulence's multiscale characteristics is essential for reliable predictions of multiphysics interactions, but remains a grand challenge even for exascale supercomputers and advanced deep learning models. The extreme-resolution data required to represent turbulence, ranging from billions to trillions of grid points, pose prohibitive computational costs for models based on architectures like vision transformers. To address this challenge, we introduce a multiscale hierarchical Turbulence Transformer that reduces sequence length from billions to a few millions and a novel RingX sequence parallelism approach that enables scalable long-context learning. We perform scaling and science runs on the Frontier supercomputer. Our approach demonstrates excellent performance up to 1.1 EFLOPS on 32,768 AMD GPUs, with a scaling efficiency of 94\%. To our knowledge, this is the first AI model for turbulence that can capture small-scale eddies down to the dissipative range in three-dimensional turbulence at high Reynolds numbers.

Yin, Junqi [ORNL] (ORCID:0000000338435520)↗

An Adaptive Middleware Framework for Scientific Computing at Extreme Scales

Large computing systems including clusters, clouds, and grids, provide high-performance capabilities that can be utilized for many applications. But as the ubiquity of these systems increases and the scope of analysis being done on them grows, there is a growing need for applications that 1) do not require users to learn the details of high performance systems, and 2) are flexible and adaptive in their usage of these systems to accommodate the best time-to-solution for end users. We introduce a new adaptive interface design and a prototype implementation within the framework of an established middleware framework, MeDICi, for high performance computing systems and describe the applicability of this adaptive design to a real-life scientific workflow. This adaptive framework provides an access model for implementing a processing pipeline using high performance systems that are not local to the data source, making it possible for the compute capabilities at one site to be applied to analysis on data being generated at another site in an automated process. This adaptive design improves overall time-to-solution by moving the data analysis task to the most appropriate resource dynamically, reacting to failures and load fluctuations.

97 MATHEMATICS AND COMPUTING↗

UNITY: Unified Memory and File Space. In: ROSS '17 Proceedings of the 7th International Workshop on Runtime and Operating Systems for Supercomputers ROSS 2017, Article No. 6

This paper describes the vision for UNITY, a new high-performance computing focused data storage abstraction that places the entire memory hierarchy, including both traditionally separated memory-and file-based data storage, into one storage continuum. Through the use of a novel API and a set of services centered around a smart runtime system, UNITY is able to provide a number of valuable and interesting benefits. The unified storage space provides a scalable and resilient data environment that dynamically manages the mapping of data onto available resources based on multiple factors, including desired persistence and energy budget considerations. By eliminating the need for high-performance computing domain scientists to develop architecture-dependent optimizations for rapidly evolving data storage technologies, UNITY addresses both ease-of-use and performance.

Jones, Terry↗

Hybrid PDES Simulation of HPC Networks Using Zombie Packets

Although high-fidelity network simulations have proven to be reliable and cost-effective tools to peer into architectural questions for high-performance computing (HPC) networks, they incur a high resource cost. The time spent in simulating a single millisecond of network traffic in the highest detail can take hours, even for static, well-behaved traffic patterns such as uniform random. Surrogate models offer a significant reduction in runtime, yet they cannot serve as complete replacements and should only be used when appropriate. Thus, there is a need for hybrid modeling, where high-fidelity simulation and surrogates run side-by-side. Here, we present a surrogate model for HPC networks in which: packets bypass the network, while the network state is left untouched, i.e., suspended. To bypass the network, we use historical data to estimate the arrival time at which every packet should be scheduled at; to suspend the network, all in-flight packets are scheduled to arrive at their destinations, and are kept in the system to awaken as zombies when switching back to high-fidelity. Speedup for a hybrid model is relative to the proportion of surrogate to high-fidelity. This light-weight surrogate obtained up to 76× speedup. Keeping the zombies in the network showed an increase in the accuracy of the high-fidelity simulation on restart when compared to restarting the network from an empty state.

HPC networks↗

Structural Simluation Toolkit (SST) v.12.0

The Structural Simulation Toolkit (SST) was developed to explore innovations in highly concurrent computing systems where the instruction set architecture (ISA), micro-architecture, and memory interact with the programming model and communications system. The package provides a fully modular design for extensive exploration of an individual system parameter as well as a parallel simulation environment based on message passing interface (MPI) which enable a high level of performance as well as the ability to look at large systems. Sandia National Laboratories is a multimission laboratory managed and operated by National Technology & Engineering Solutions of Sandia, LLC, a wholly owned subsidiary of Honeywell International Inc., for the U.S. Department of Energy’s National Nuclear Security Administration under contract DE-NA0003525.

Rodrigues, ArunF.↗

Understanding power and energy utilization in large scale production physics simulation codes

Power is an often-cited reason for the move to advanced architectures on the path to Exascale computing. Here, this is due to practical considerations related to delivering enough power to successfully site and operate these machines, as well as concerns about energy usage while running large simulations. Since obtaining accurate power measurements can be challenging, it may be tempting to use the processor thermal design power (TDP) as a surrogate due to its simplicity and availability. However, TDP is not indicative of typical power usage while running simulations. Using commodity and advanced technology systems at Lawrence Livermore and Sandia National Labs, we performed a series of experiments to measure power and energy usage in running simulation codes. These experiments indicate that large scale Lawrence Livermore simulation codes are significantly more efficient than a simple processor TDP model might suggest.

HPC↗

SD-WAN Evaluation Criteria for a Defense Information Systems Network Expeditionary Customer Edge

The Defense Information Systems Agency (DISA) has identified a service provision gap midway between the capacity and capability of a Defense Information Systems Network (DISN) transport Edge Points of Presence (POP) and U.S. Department of Defense (DoD) Enterprise Classified Travel Kit (DECTK). Some deployed force 100-user Tactical Operations Centers are in field locations far removed from the DISN transport core, in challenging environmental conditions with limitations on space, weight, power, and cooling for network equipment. As part of a multi-phase project, Pacific Northwest National Laboratory (PNNL) will gather and analyze requirements, design, develop, prototype, and test a miniaturized and ruggedized DISN Customer Edge POP scaled to support approximately 100 users in a field Tactical Operations Center. An Expeditionary Customer Edge (ECE) will be more suitable for field deployment than a DISN transport Edge POP, using ruggedized hardware and network function virtualization (NFV) to operate in challenging field environmental conditions, plus reduce weight, power, and cooling requirements. A key enabling technology for ECE is Software Defined Wide Area Networking (SD-WAN). This document provides: • A brief overview of SD-WAN use cases and how they apply to ECE • How SD-WAN technology compares to existing networks like DISN that use Optical Transport Networking (OTN) and Multiprotocol Label Switching (MPLS) • SD-WAN functional requirements relevant to ECE • A description of an ECE Virtual Prototype, including simulated wide area network paths and a DISN-representative implementation of MPLS, in Cisco Modeling Labs • Detailed network flow walkthroughs • Evaluation criteria based on ECE-relevant SD-WAN functional requirements Finally, an appendix provides an informal evaluation of Speedify—a commercial retail Virtual Private Network service—against the ECE SD-WAN evaluation criteria.

42 ENGINEERING↗

Understanding Power and Energy Utilization in Large Scale Production Physics Simulation Codes

Power is an often-cited reason for moving to advanced architectures on the path to Exascale computing. This is due to the practical concern of delivering enough power to successfully site and operate these machines, as well as concerns over energy usage while running large simulations. Since accurate power measurements can be difficult to obtain, processor thermal design power (TDP) is a possible surrogate due to its simplicity and availability. However, TDP is not indicative of typical power usage while running simulations. Using commodity and advance technology systems at Lawrence Livermore National Laboratory (LLNL) and Sandia National Laboratory, we performed a series of experiments to measure power and energy usage in running simulation codes. These experiments indicate that large scale LLNL simulation codes are significantly more efficient than a simple processor TDP model might suggest.

97 MATHEMATICS AND COMPUTING↗

Ps and Qs: Quantization-Aware Pruning for Efficient Low Latency Neural Network Inference

Efficient machine learning implementations optimized for inference in hardware have wide-ranging benefits, depending on the application, from lower inference latency to higher data throughput and reduced energy consumption. Two popular techniques for reducing computation in neural networks are pruning, removing insignificant synapses, and quantization, reducing the precision of the calculations. In this work, we explore the interplay between pruning and quantization during the training of neural networks for ultra low latency applications targeting high energy physics use cases. Techniques developed for this study have potential applications across many other domains. We study various configurations of pruning during quantization-aware training, which we term quantization-aware pruning, and the effect of techniques like regularization, batch normalization, and different pruning schemes on performance, computational complexity, and information content metrics. We find that quantization-aware pruning yields more computationally efficient models than either pruning or quantization alone for our task. Further, quantization-aware pruning typically performs similar to or better in terms of computational efficiency compared to other neural architecture search techniques like Bayesian optimization. Surprisingly, while networks with different training configurations can have similar performance for the benchmark application, the information content in the network can vary significantly, affecting its generalizability.

46 INSTRUMENTATION RELATED TO NUCLEAR SCIENCE AND ↗

Communication Simulations for Power System Applications

New smart grid technologies and concepts, such as dynamic pricing, demand response, dynamic state estimation, and wide area monitoring, protection, and control, are expected to require considerable communication resources. As the cost of retrofit can be high, future power grids will require the integration of high-speed, secure connections with legacy communication systems, while still providing adequate system control and security. While considerable work has been performed to create co-simulators for the power domain with load models and market operations, limited work has been performed in integrating communications directly into a power domain solver. The simulation of communication and power systems will become more important as the two systems become more inter-related. This paper will discuss ongoing work at Pacific Northwest National Laboratory to create a flexible, high-speed power and communication system co-simulator for smart grid applications. The framework for the software will be described, including architecture considerations for modular, high performance computing and large-scale scalability (serialization, load balancing, partitioning, cross-platform support, etc.). The current simulator supports the ns-3 (telecommunications) and GridLAB-D (distribution systems) simulators. Ongoing and future work will be described, including planned future expansions for a traditional transmission solver. A test case using the co-simulator, utilizing a transactive demand response system created for the Olympic Peninsula and AEP gridSMART demonstrations, requiring two-way communication between distributed and centralized market devices, will be used to demonstrate the value and intended purpose of the co-simulation environment.

co-simulation↗

UNITY: Unified Memory and File Space

This paper describes the vision for UNITY, a new high-performance computing focused data storage abstraction that places the entire memory hierarchy, including both traditionally separated memoryand file-based data storage, into one storage continuum. Through the use of a novel API and a set of services centered around a smart runtime system, UNITY is able to provide a number of valuable and interesting benefits. The unified storage space provides a scalable and resilient data environment that dynamically manages the mapping of data onto available resources based on multiple factors, including desired persistence and energy budget considerations. By eliminating the need for high-performance computing domain scientists to develop architecture-dependent optimizations for rapidly evolving data storage technologies, UNITY addresses both ease-of-use and performance.

Otstott, Douglas↗

Quantum Information Processing in the Frequency Domain

Quantum technologies promise to revolutionize information processing through dramatic computational speed-ups and ultimate security compared with classical counterparts. As matter-based qubit systems increase in size, however, counteracting decoherence becomes challenging and places limits on how large a stand-alone system can be. These limitations, coupled with the many quantum computing platforms under development, have led researchers to view modular architectures, connected by a quantum network, as a promising path to large-scale quantum computing. Our group reported in several studies this year, progress on multiple fronts in such systems, using photons as quantum information carriers in the frequency domain.

97 MATHEMATICS AND COMPUTING↗

Studying Performance Portability of LAMMPS across Diverse GPU-based Platforms

The molecular dynamics simulation software, LAMMPS, utilizes the Kokkos acceleration library to port computation to a diverse set of architectures including those based on GPU accelerators. In addition to Kokkos, LAMMPS contains a vast code base that leverages the CUDA application programming interface using library functions such as cuFFT, CUDA’s fast-fourier transform (FFT) library, and, more recently, also support for AMD’s Heterogeneous Interface for Portability (HIP) that is rapidly growing. While preparing LAMMPS tests for the AMD GPU-based test system precursors to Frontier, we investigated several strategies for accelerating LAMMPS on AMD GPUs, using the AMD Instinct MI100 and MI250X. In this work, we integrated the HIP FFT library, hipFFT, into the particle-particle particle-mesh (PPPM) long-range solver, which allowed the porting of PPPM calculations to the GPUs. Kokkos behavior on the MI100 and MI250X was also investigated through the package kokkos command of LAMMPS, targeting com- munication, memory usage, and particle grid decomposition. The Tersoff, Reax, Lennard-Jones (LJ), EAM, Granular, and PPPM potentials were investigated in this effort, and results from these experiments are provided. The selected potentials were run on Spock (AMD Instinct MI100), Crusher (AMD Instinct MI250X), AFW HPC11 (NVIDIA A100) and Summit (NVIDIA V100), for comparison. Operational roofline models were constructed and analyzed for the Tersoff, Reax, and Lennard-Jones potentials on Crusher and Summit.

Hagerty, Nick↗