Search NASA⌕ Search

SEARCH · Search NASA

Results for “Memory Optimization”

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 613 records · Page 34
All sourcesNASA NTRSDOE OSTI
DOE OSTI2024

Exploring code portability solutions for HEP with a particle tracking test code

Traditionally, high energy physics (HEP) experiments have relied on x86 CPUs for the majority of their significant computing needs. As the field looks ahead to the next generation of experiments such as DUNE and the High-Luminosity LHC, the computing demands are expected to increase dramatically. To cope with this increase, it will be necessary to take advantage of all available computing resources, including GPUs from different vendors. A broad landscape of code portability tools—including compiler pragma-based approaches, abstraction libraries, and other tools—allow the same source code to run efficiently on multiple architectures. In this paper, we use a test code taken from a HEP tracking algorithm to compare the performance and experience of implementing different portability solutions. While in several cases portable implementations perform close to the reference code version, we find that the performance varies significantly depending on the details of the implementation. Achieving optimal performance is not easy, even for relatively simple applications such as the test codes considered in this work. Several factors can affect the performance, such as the choice of the memory layout, the memory pinning strategy, and the compiler used. The compilers and tools are being actively developed, so future developments may be critical for their deployment in HEP experiments.

72 PHYSICS OF ELEMENTARY PARTICLES AND FIELDS↗
DOE OSTI2024

Performance Study of CXL Memory Topology

This paper presents a comprehensive evaluation of the performance impact of various Compute Express Link (CXL) memory topologies, with a particular emphasis on CXL switches, in the context of High- Performance Computing (HPC) and Large Language Model (LLM) inference workloads. Our study unveils significant performance variations across different topologies, demonstrating that certain configurations yield superior performance for specific workloads. These findings underscore the critical importance of tailored topol- ogy selection in optimizing system performance. Additionally, we address the inherent challenges associated with integrating CXL switches, including overhead considerations and routing complex- ities. Our research highlights the necessity for thorough evalua- tion methodologies to fully leverage CXL technology’s potential in contemporary computing environments. These insights provide valuable guidance for system architects and data center operators in designing and optimizing CXL-based infrastructures for diverse workload requirements.

CXL, memory, Artificial Intelligence (AI), HPC↗
DOE OSTI2022

Accurate Compression of Tabulated Chemistry Models with Partition of Unity Networks

Tabulated chemistry models are widely used to simulate large-scale turbulent fires in applications including energy generation and fire safety. Tabulation via piecewise Cartesian interpolation suffers from the curse-of-dimensionality, leading to a prohibitive exponential growth in parameters and memory usage as more dimensions are considered. Artificial neural networks (ANNs) have attracted attention for constructing surrogates for chemistry models due to their ability to perform high-dimensional approximation. However, due to well-known pathologies regarding the realization of suboptimal local minima during training, in practice they do not converge and provide unreliable accuracy. Partition of unity networks (POUnets) are a recently introduced family of ANNs which preserve notions of convergence while performing high-dimensional approximation, discovering a mesh-free partition of space which may be used to perform optimal polynomial approximation. In this work, we assess their performance with respect to accuracy and model complexity in reconstructing unstructured flamelet data representative of nonadiabatic pool fire models. Our results show that POUnets can provide the desirable accuracy of classical spline-based interpolants with the low memory footprint of traditional ANNs while converging faster to significantly lower errors than ANNs. For example, we observe POUnets obtaining target accuracies in two dimensions with 40 to 50 times less memory and roughly double the compression in three dimensions. We also address the practical matter of efficiently training accurate POUnets by studying convergence over key hyperparameters, the impact of partition/basis formulation, and the sensitivity to initialization.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2021

Optimization of Asynchronous Communication Operations through Eager Notifications

UPC++ is a C++ library implementing the Asynchronous Partitioned Global Address Space (APGAS) model. We propose an enhancement to the completion mechanisms of UPC++ used to synchronize communication operations that is designed to reduce overhead for on-node operations. Our enhancement permits eager delivery of completion notification in cases where the data transfer semantics of an operation happen to complete synchronously, for example due to the use of shared-memory bypass. This semantic relaxation allows removing significant overhead from the critical path of the implementation in such cases. We evaluate our results on three different representative systems using a combination of microbenchmarks and five variations of the the HPCChallenge RandomAccess benchmark implemented in UPC++ and run on a single node to accentuate the impact of locality. We find that in RMA versions of the benchmark written in a straightforward manner (without manually optimizing for locality), the new eager notification mode can provide up to a 25% speedup when synchronizing with promises and up to a 13.5x speedup when synchronizing with conjoined futures. We also evaluate our results using a graph matching application written with UPC++ RMA communication, where we measure overall speedups of as much as 11% in single-node runs of the unmodified application code, due to our transparent enhancements.

Kamil, Amir↗
DOE OSTI2014

A nonvoxel-based dose convolution/superposition algorithm optimized for scalable GPU architectures

Purpose: Real-time adaptive planning and treatment has been infeasible due in part to its high computational complexity. There have been many recent efforts to utilize graphics processing units (GPUs) to accelerate the computational performance and dose accuracy in radiation therapy. Data structure and memory access patterns are the key GPU factors that determine the computational performance and accuracy. In this paper, the authors present a nonvoxel-based (NVB) approach to maximize computational and memory access efficiency and throughput on the GPU. Methods: The proposed algorithm employs a ray-tracing mechanism to restructure the 3D data sets computed from the CT anatomy into a nonvoxel-based framework. In a process that takes only a few milliseconds of computing time, the algorithm restructured the data sets by ray-tracing through precalculated CT volumes to realign the coordinate system along the convolution direction, as defined by zenithal and azimuthal angles. During the ray-tracing step, the data were resampled according to radial sampling and parallel ray-spacing parameters making the algorithm independent of the original CT resolution. The nonvoxel-based algorithm presented in this paper also demonstrated a trade-off in computational performance and dose accuracy for different coordinate system configurations. In order to find the best balance between the computed speedup and the accuracy, the authors employed an exhaustive parameter search on all sampling parameters that defined the coordinate system configuration: zenithal, azimuthal, and radial sampling of the convolution algorithm, as well as the parallel ray spacing during ray tracing. The angular sampling parameters were varied between 4 and 48 discrete angles, while both radial sampling and parallel ray spacing were varied from 0.5 to 10 mm. The gamma distribution analysis method (γ) was used to compare the dose distributions using 2% and 2 mm dose difference and distance-to-agreement criteria, respectively. Accuracy was investigated using three distinct phantoms with varied geometries and heterogeneities and on a series of 14 segmented lung CT data sets. Performance gains were calculated using three 256 mm cube homogenous water phantoms, with isotropic voxel dimensions of 1, 2, and 4 mm. Results: The nonvoxel-based GPU algorithm was independent of the data size and provided significant computational gains over the CPU algorithm for large CT data sizes. The parameter search analysis also showed that the ray combination of 8 zenithal and 8 azimuthal angles along with 1 mm radial sampling and 2 mm parallel ray spacing maintained dose accuracy with greater than 99% of voxels passing the γ test. Combining the acceleration obtained from GPU parallelization with the sampling optimization, the authors achieved a total performance improvement factor of >175 000 when compared to our voxel-based ground truth CPU benchmark and a factor of 20 compared with a voxel-based GPU dose convolution method. Conclusions: The nonvoxel-based convolution method yielded substantial performance improvements over a generic GPU implementation, while maintaining accuracy as compared to a CPU computed ground truth dose distribution. Such an algorithm can be a key contribution toward developing tools for adaptive radiation therapy systems.

60 APPLIED LIFE SCIENCES↗
DOE OSTI2024

Acceleration of Graph Neural Network-Based Prediction Models in Chemistry via Co-Design Optimization on Intelligence Processing Units

Atomic structure prediction and associated property calculations are the bedrock of chemical physics. Since high-fidelity ab initio modeling techniques for computing the structure and properties can be prohibitively expensive, this motivates the development of machine-learning (ML) models that make these predictions more efficiently. Training graph neural networks over large atomistic databases introduces unique computational challenges such as the need to process millions of small graphs with variable size and support communication patterns that are distinct from learning over large graphs such as social networks. We demonstrate a novel hardware-software co-design approach to scale up the training of atomistic graph neural networks (GNN) for structure and property prediction. First, to eliminate redundant computation and memory associated with alternative padding techniques and to improve throughput via minimizing communication, we formulate the effective coalescing of the batches of variable-size atomistic graphs as the bin packing problem and introduce a hardware-agnostic algorithm to pack these batches. In addition, we propose hardware-specific optimizations including a planner and vectorization for the gather-scatter operations targeted for Graphcore’s Intelligence Processing Unit (IPU), as well as model-specific optimizations such as merged communication collectives and optimized softplus. Putting these all together, we demonstrate the effectiveness of the proposed co-design approach by providing an implementation of a well-established atomistic GNN on the Graphcore IPUs. We evaluate the training performance on multiple atomistic graph databases with varying degrees of graph counts, sizes and sparsity. Here, we demonstrate that such a co-design approach can reduce the training time of atomistic GNNs and can improve the performance by up to 1.5× compared to the baseline implementation of the model on the IPUs. Additionally, we compare our IPU implementation with a Nvidia GPU-based implementation and show that our atomistic GNN implementation on the IPUs can run 1.8× faster on average compared to the execution time on the GPUs.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗
DOE OSTI2025

Transfer Learning Trained LSTM Models for Household Load Profile Forecasting

Grid edge renewable energy resources, such as rooftop solar photovoltaics, closely interact with consumer load profiles. Therefore, forecasting future electricity demand, ideally at the individual household level, is indispensable. In this paper, we present a transfer learning enhanced household load profile forecasting method. First, we tune a long short-term memory forecasting model to perform day-ahead prediction of household electricity load profiles. Then we improve these individualized models using transfer learning, and we use k-means clustering to create optimal source data sets. We find average improvements of 4.38% (largest improvement of 10.71%) when the entire data set was used to train the source model and 2.45% (largest improvement of 11.57%) in the mean absolute error when households were first clustered and used to train separate source models for each cluster. We find that transfer learning with clustered data can effectively boost the forecasting performance of the LSTM models. We use realistic household power measurements for 148 real residential households in Austin, Texas.

deep learning↗
DOE OSTI2021

Implementing Arbitrary/Common Concurrent Writes of CRCW PRAM

The Parallel Random Access Machines (PRAM) abstraction is the simplest and most elegant algorithmic model for the design and analysis of parallel algorithms. It consists of different models categorized based on the underlying memory access mode used, the most powerful of which is the Concurrent Read Concurrent Write (CRCW) model. A PRAM algorithm describes a series of rounds, each of which consists of a collection of operations that can be executed concurrently within the same time step. However, the lack of support for concurrent memory accesses and the prevalence of asynchronous programming models led to the belief that implementing CRCW PRAM algorithms is unattainable and prompted many to avoid this model except for theoretical studies of optimal performance.In this work, we study the arbitrary and common concurrent writes in the CRCW PRAM model and explore implementation challenges on general-purpose systems. Moreover, we examine current practices for implementing common/arbitrary concurrent writes and propose a new efficient lightweight and thread-safe method to implement concurrent writes through leveraging atomic instructions. To demonstrate the efficacy of our method, we developed OpenMP kernels for classical CRCW PRAM algorithms and provide experimental results and comparisons based on run time performance measured over the x86 multicore architecture. Our results show a performance speedup compared to current practices up to 4.5x across all our benchmarks.

Ghanim, Fady↗
DOE OSTI2015

SU-E-I-54: Effective Dose and Radiation Cancer Risks for Scoliosis Patients Undergoing Full Spine Radiography

Purpose: Scoliotic patients underwent a lot of radiologic examinations during the control and treatment periods. This study used the PCXMC program to calculate the effective dose of the patients and assess the radiation cancer risks. Methods: Seventy five scoliotic patients were examined using CR or DR systems during the control and treatment periods in Chang Gung Memorial Hospital. The technical factors were recorded for each patient during his/her control and treatment period. The entrance surface dose was measured using thermoluminence dosimeters and derived from technical factors and irradiated geometry. The effective dose of patients and relative radiation cancer risks were calculated by the PCXMC program. All required information regarding patient age and sex, the x-ray spectra, and the tube voltage and current were registered. The radiation risk were estimated using the model developed by the BEIR VII committee (2006). Results: The effective doses of full spine radiography with anteroposterior and lateral projections were 0.626 mSv for patients using DR systems, and 0.483mSv for patients using CR systems, respectively. The dose using DR system was 29.6% higher than those using CR system. The maximum organ dose was observed in the breast for both projections in all the systems. The risk of exposure—induced cancer death (REID) of patients for DR and CR systems were 0.009% and 0.007%, respectively. Conclusion: The risk estimates were regarded with healthy skepticism, placed more emphasis on the magnitude of the risk. The effective doses estimated in this study could be served as a reference for radiologists and technologists and demonstrate the necessity to optimize patient protection for full spine radiography though the effective doses are not at the level to induce deterministic effects and not significant in the stochastic effect. This study was supported by the grants from the Chang Gung Memorial Hospital (CMRPD1D0421)

60 APPLIED LIFE SCIENCES↗
DOE OSTI2013

TOPS2 - Towards Optimal Petascale Simulations

Our goal in this project was to design scalable numerical algorithms needed by SciDAC applications that adapt to use evolving hardware resources as efficiently as possible. Our primary challenge is minimizing communication costs, where communication means moving data either between levels of a memory hierarchy (L1 cache to L2 cache to main memory etc.) or between processors over a network. Floating point rates are improving exponentially faster than bandwidth, which is improving exponentially faster than latency. So our goal is to minimize communication. We describe our progress in this area, both for direct and iterative linear algebra. In both areas we have (1) identified lower bounds on the amount of communication (measured both by the number of words moved and the number of messages) required to perform these algorithms, (2) analyzed existing algorithms, which by and large do not attain these lower bounds, and (3) identified or invented new algorithms that do attain them, and evaluated their speedups, which can be quite large.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2025

Erbium quantum memory platform with long optical coherence via back-end-of-line deposition on foundry-fabricated photonics

Realizing scalable quantum interconnects necessitates the integration of solid-state quantum memories with foundry photonics processes. While prior photonic integration efforts have relied upon specialized, laboratory-scale fabrication techniques, this work demonstrates the monolithic integration of a quantum memory platform with low-loss foundry photonic circuits via back-end-of-line deposition. We deposited thin films of titanium dioxide (Ti⁢O 2 ) doped with erbium (Er) onto silicon nitride nanophotonic waveguides and studied Er optical coherence at subkelvin temperatures with photon echo techniques. Furthermore, we suppressed optical dephasing through ex-situ oxygen annealing and optimized measurement conditions, which yielded an optical coherence time of 64 μ⁢s (a 5-kHz homogeneous linewidth) and slow spectral diffusion of 27 kHz over 4 ms, results that are comparable to state-of-the-art Er nanophotonic devices. Combined with second-long electron spin lifetimes and demonstrated electrical control of Er emission, our findings establish Er:Ti⁢O 2 on foundry photonics as a manufacturable platform for ensemble and single-ion quantum memories.

Electro-optic effects↗
DOE OSTI2014

Effects of Isophorone Diisocyanate on the Thermal and Mechanical Properties of Shape-Memory Polyurethane Foams

Previously developed shape-memory polymer foams display fast actuation in water due to plasticization of the polymer network. The actuation presents itself as a depression in the glass-transition temperature when moving from dry to aqueous conditions; this effect limits the working time of the foam to 10 min when used in a transcatheter embolic device. Reproducible foams are developed by altering the chemical backbone, which can achieve working times of greater than 20 min. This is accomplished by incorporating isophorone diisocyanate into the foam, resulting in increased hydrophobicity, glass transitions, and actuation time. Furthermore, this delayed actuation, when compared with previous systems, allows for more optimal working time in clinical applications.

36 MATERIALS SCIENCE↗
DOE OSTI2026

Efficient derivative computation for unsteady fatigue-constrained nonlinear aero-structural wind turbine blade optimization

Gradient-based optimization offers significant efficiency advantages for wind turbine blade design, but its application has often been limited by the cost and accuracy of finite-difference derivative calculations, especially when fatigue constraints are considered. In this work, we systematically compare and evaluate four differentiation techniques, namely algorithmic differentiation, implicit differentiation, sparsity exploitation, and parallelization, to determine their effectiveness in computing accurate gradients through time-domain aero-structural simulations. By integrating these techniques with unsteady nonlinear aerodynamic and structural models, we develop software designed for accurate gradient computation. We show that combining these techniques addresses memory and runtime challenges associated with long simulations required by design load cases. Specifically, the most effective combination reduces derivative computation wall time by over an order of magnitude compared to finite differencing while maintaining superior accuracy. We demonstrate this approach in a proof-of-concept aero-structural optimization of a wind turbine blade that improves the cost of energy by 12.78 %. This comparative study establishes a viable approach for fatigue-aware blade design that balances computational efficiency with modeling accuracy.

17 WIND ENERGY↗
DOE OSTI2015

Thread-level parallelization and optimization of NWChem for the Intel MIC architecture

In the multicore era it was possible to exploit the increase in on-chip parallelism by simply running multiple MPI processes per chip. Unfortunately, manycore processors' greatly increased thread- and data-level parallelism coupled with a reduced memory capacity demand an altogether different approach. In this paper we explore augmenting two NWChem modules, triples correction of the CCSD(T) and Fock matrix construction, with OpenMP in order that they might run efficiently on future manycore architectures. As the next NERSC machine will be a self-hosted Intel MIC (Xeon Phi) based supercomputer, we leverage an existing MIC testbed at NERSC to evaluate our experiments. In order to proxy the fact that future MIC machines will not have a host processor, we run all of our experiments in native mode. We found that while straightforward application of OpenMP to the deep loop nests associated with the tensor contractions of CCSD(T) was sufficient in attaining high performance, significant e ort was required to safely and efeciently thread the TEXAS integral package when constructing the Fock matrix. Ultimately, our new MPI+OpenMP hybrid implementations attain up to 65× better performance for the triples part of the CCSD(T) due in large part to the fact that the limited on-card memory limits the existing MPI implementation to a single process per card. Additionally, we obtain up to 1.6× better performance on Fock matrix constructions when compared with the best MPI implementations running multiple processes per card.

CCSD(T)↗
DOE OSTI2023

Code modernization strategies for short-range non-bonded molecular dynamics simulations

Modern HPC systems are increasingly relying on greater core counts and wider vector registers. Thus, applications need to be adapted to fully utilize these hardware capabilities. One class of applications that can benefit from this increase in parallelism are molecular dynamics simulations. In this paper, we describe our efforts at modernizing the ESPResSo++ simulation package for molecular dynamics by restructuring its particle data layout for efficient memory accesses and applying vectorization techniques to benefit the calculation of short-range non-bonded forces, which results in an overall three times speedup and serves as a baseline for further optimizations. We also implement fine-grained parallelism for multi-core CPUs through HPX, a C++ runtime system which uses lightweight threads and an asynchronous many-task approach to maximize concurrency. Our goal is to evaluate the performance of an HPX-based approach compared to the bulk-synchronous MPI-based implementation. This requires the introduction of an additional layer to the domain decomposition scheme that defines the task granularity. On spatially inhomogeneous systems, which impose a corresponding load-imbalance in traditional MPI-based approaches, we demonstrate that by choosing an optimal task size, the efficient work-stealing mechanisms of HPX can overcome the overhead of communication resulting in an overall 1.4 times speedup compared to the baseline MPI version.

97 MATHEMATICS AND COMPUTING↗
DOE OSTI2024

Fracture Characterization Via AI‐Assisted Analysis of Temperature Logs

Abstract Fractures control fluid flow, mass transport, and heat transfer in a geothermal reservoir. This makes accurate characterization of fracture networks a prerequisite for optimal design and control of a reservoir's exploitation. We develop a deep‐learning procedure to identify fracture locations via interpretation of temporally and spatially continuous downhole temperature measurements. A long short‐term memory fully convolutional network (LSTM‐FCN) is used both to capture long‐term dependencies in sequential temperature data and to distill local features around fractures. A wellbore and fractured‐reservoir thermal model is established to generate temperature data for network training. The trained LSTM‐FCN exhibits a unique ability to detect multiple fractures intersecting a borehole. We use the LSTM‐FCN algorithm to evaluate the effectiveness of different‐stage wellbore temperature measurements on fracture detection in a complex fractured system. Our experiments reveal that the use of various‐stage temperature information as an input feature set improves the robustness of fracture detection to noise interference. This study indicates the practical feasibility of obtaining accurate fracture‐network reconstructions from temperature signals, at reasonable computational cost.

Yang, Xiaoyu↗
DOE OSTI2024

MassiveGNN: Efficient Training via Prefetching for Massively Connected Distributed Graphs

Graph Neural Networks (GNN) are indispensable in learning from graph-structured data, yet their rising computational costs, especially on massively connected graphs, pose significant challenges in terms of execution performance. To tackle this, distributed-memory solutions such as partitioning the graph to concurrently train multiple replicas of GNNs are in practice. However, approaches requiring a partitioned graph usually suffer from communication overhead and load imbalance, even under optimal partitioning and communication strategies due to irregularities in the neighborhood minibatch sampling. This paper proposes practical trade-offs for improving the sampling and communication overheads for representation learn- ing on distributed graphs (using popular GraphSAGE architecture) by developing a parameterized prefetch and eviction scheme on top of the state-of-the-art Amazon DistDGL distributed GNN framework, demonstrating about 15–40% improvement in end-to-end training performance on the NERSC Perlmutter supercomputer for various OGB datasets.

Machine Leanring, high performance comptuing, grap↗
DOE OSTI2023

A Hybrid Optimization and Deep Learning Algorithm for Cyber-Resilient DER Control

With the proliferation of distributed energy resources (DERs) in the distribution grid, it is a challenge to effectively control a large number of DERs resilient to the communication and security disruptions, as well as to provide the online grid services, such as voltage regulation and virtual power plant (VPP) dispatch. To this end, a hybrid feedback-based optimization algorithm along with deep learning forecasting technique is proposed to specifically address the cyber-related issues. The online decentralized feedback-based DER optimization control requires timely, accurate voltage measurement from the grid. However, in practice such information may not be received by the control center or even be corrupted. Therefore, the long short-term memory (LSTM) deep learning algorithm is employed to forecast delayed/missed/attacked messages with high accuracy. The IEEE 37-node feeder with high penetration of PV systems is used to validate the efficiency of the proposed hybrid algorithm. The results show that 1) the LSTM-forecasted lost voltage can effectively improve the performance of the DER control algorithm in the practical cyber-physical architecture; and 2) the LSTM forecasting strategy outperforms other strategies of using previous message and skipping dual parameter update.

cyber-resilient algorithm↗