Engineering Papers⌕ Search

SEARCH · Engineering Papers

Results for “Memory benchmark”

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 289 records · Page 16

Challenges and Opportunities of User-Level File Systemsfor HPC

The performance gap between magnetic disks and data processing on HPC systems has become that huge that an efficient data processing can only be achieved by introducing non-volatile memory (NVRAM) as a new storage tier. Although the benefits of hierarchical storage have been adequately demonstrated to the point that the newest leadership class HPC systems will employ burst buffers, critical questions remain for supporting hierarchical storage systems, including: How should we present hierarchical storage systems to user applications, such that they are easy to use and that application code is portable across systems? How should we manage data movement through a storage hierarchy for best performance and resilience of data? How do the particular I/O use cases mandate the way we manage data? There have been many efforts to explore this space in the form of file systems, with increasingly more implemented at the user level. This is because it is relatively easy to swap in new, specialized user-level file systems for use by applications on a case-by-case basis, as opposed to the current mainstream approach of using general-purpose, system-level file systems which may not be optimized for HPC workloads and must be installed by administrators. In contrast, file systems at the user level can be tailored for specific HPC workloads for high performance and can be used by applications without administrator intervention. Many user-level file system developers have found themselves “having to reinvent the wheel” to implement various optimizations in their file systems. Thus, a main goal of this meeting was to bring together experts in I/O performance, file systems, and storage, and collectively explore the space of current and future problems and solutions for I/O on hierarchical storage systems in order to begin a community effort in enabling user-level file system support for HPC systems. We had a lively week of learning about each other’s approaches as well as unique I/O use cases that can influence the design of a community-driven file and storage system standards. The agenda for this meeting contained talks from participants on the following high level topics: HPC storage and I/O support today; what do HPC users need for I/O; existing userlevel file system efforts; object stores and other alternative storage systems; and components for building user-level file systems. The talks were short and intended to be conversation starters for more in-depth discussions with the whole group. The participants engaged in lengthy discussions on various questions that arose from the talks including: Are we ready to program to a memory hierarchy versus block devices? Are the needs of HPC users reflected in our existing file systems and storage systems? Should we drop or keep POSIX moving forward? What do we mean when we say “user-level file system”? Do we all mean the same thing? How should the IO 500 benchmark be defined so it is fair and useful? and How are stage-in and stage-out actually going to work? The report for this seminar contains a record of the talks from the participants as well as the resulting discussions. Our hope is that the effort initiated during this seminar will result in long-term collaborations that will benefit the HPC community as a whole.

97 MATHEMATICS AND COMPUTING↗

Evaluating Online Global Recovery with Fenix Using Application-Aware In-Memory Checkpointing Techniques

Exascale systems promise the potential for computation at unprecedented scales and resolutions, but achieving exascale by the end of this decade presents significant challenges. A key challenge is due to the very large number of cores and components and the resulting mean time between failures (MTBF) in the order of hours or minutes. Since the typical run times of target scientific applications are longer than this MTBF, fault tolerance techniques will be essential. An important class of failures that must be addressed is process or node failures. While checkpoint/restart (C/R) is currently the most widely accepted technique for addressing processor failures, coordinated, stable-storage-based global C/R might be unfeasible at exascale when the time to checkpoint exceeds the expected MTBF. This paper explores transparent recovery via implicitly coordinated, diskless, application-driven checkpointing as a way to tolerate process failures in MPI applications at exascale. The discussed approach leverages User Level Failure Mitigation (ULFM), which is being proposed as an MPI extension to allow applications to create policies for tolerating process failures. Specifically, this paper demonstrates how different implementations of application-driven in-memory checkpoint storage and recovery compare in terms of performance and scalability. We also experimentally evaluate the effectiveness and scalability of the Fenix online global recovery framework on a production system -the Titan Cray XK7 at ORNL- and demonstrate the ability of Fenix to tolerate dynamically injected failures using the execution of four benchmarks and mini-applications with different behaviors.

97 MATHEMATICS AND COMPUTING↗

Comparing Server Energy Use and Efficiency Using Small Sample Sizes

This report documents a demonstration that compared the energy consumption and efficiency of a limited sample size of server-type IT equipment from different manufacturers by measuring power at the server power supply power cords. The results are specific to the equipment and methods used. However, it is hoped that those responsible for IT equipment selection can used the methods described to choose models that optimize energy use efficiency. The demonstration was conducted in a data center at Lawrence Berkeley National Laboratory in Berkeley, California. It was performed with five servers of similar mechanical and electronic specifications; three from Intel and one each from Dell and Supermicro. Server IT equipment is constructed using commodity components, server manufacturer-designed assemblies, and control systems. Server compute efficiency is constrained by the commodity component specifications and integration requirements. The design freedom, outside of the commodity component constraints, provides room for the manufacturer to offer a product with competitive efficiency that meets market needs at a compelling price. A goal of the demonstration was to compare and quantify the server efficiency for three different brands. The efficiency is defined as the average compute rate (computations per unit of time) divided by the average energy consumption rate. The research team used an industry standard benchmark software package to provide a repeatable software load to obtain the compute rate and provide a variety of power consumption levels. Energy use when the servers were in an idle state (not providing computing work) were also measured. At high server compute loads, all brands, using the same key components (processors and memory), had similar results; therefore, from these results, it could not be concluded that one brand is more efficient than the other brands. The test results show that the power consumption variability caused by the key components as a group is similar to all other components as a group. However, some differences were observed. The Supermicro server used 27 percent more power at idle compared to the other brands. The Intel server had a power supply control feature called cold redundancy, and the data suggest that cold redundancy can provide energy savings at low power levels. Test and evaluation methods that might be used by others having limited resources for IT equipment evaluation are explained in the report.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

$\mathrm{PPT}$-Multicore: performance prediction of Open$\mathrm{MP}$ applications using reuse profiles and analytical modeling

In this report we present PPT-Multicore, an analytical model embedded in the Performance Prediction Toolkit (PPT) to predict parallel applications’ performance running on a multicore processor. PPT-Multicore builds upon our previous work towards a multicore cache model. We extract LLVM basic block labeled memory trace using an architecture-independent LLVM-based instrumentation tool only once in an application’s lifetime. The model uses the memory trace and other parameters from an instrumented sequentially executed binary. We use probabilistic and computationally efficient reuse profiles to predict the cache hit rates and runtimes of OpenMP programs’ parallel sections. We model Intel’s Broadwell, Haswell, and AMD’s Zen2 architectures and validate our framework using different applications from PolyBench and PARSEC benchmark suites. The results show that PPT-Multicore can predict cache hit rates with an overall average error rate of 1.23% while predicting the runtime with an error rate of 9.08%.

97 MATHEMATICS AND COMPUTING↗

BlackjackBench: Portable Hardware Characterization

DARPA's AACE project aimed to develop Architecture Aware Compiler Environments that automatically characterize the hardware and optimize the application codes accordingly. We present the BlackjackBench suite, a collection of portable micro-benchmarks that automate system characterization, plus statistical analysis techniques for interpreting the results. BlackjackBench discovers the effective sizes and speeds of the hardware environment rather than the often unattainable peak values. We aim at hardware features that can be observed by running executables generated by existing compilers from standard C codes. We characterize the memory hierarchy, including cache sharing and NUMA characteristics of the system, properties of the processing cores affecting execution speed, and the length of the OS scheduler time slot. We show how these features of modern multicores can be discovered programmatically. We also show how the features could interfere with each other resulting in incorrect interpretation of the results, and how established classification and statistical analysis techniques reduce experimental noise and aid automatic interpretation of results.

99 GENERAL AND MISCELLANEOUS↗

Correlation energy extrapolation by many-body expansion

Accounting for electron correlation is required for high accuracy calculations of molecular energies. The full configuration interaction (CI) approach can fully capture the electron correlation within a given basis, but it does so at a computational expense that is impractical for all but the smallest chemical systems. In this work, a new methodology is presented to approximate configuration interaction calculations at a reduced computational expense and memory requirement, namely, the correlation energy extrapolation by many-body expansion (CEEMBE). This method combines a MBE approximation of the CI energy with an extrapolated correction obtained from CI calculations using subsets of the virtual orbitals. The extrapolation approach is inspired by, and analogous to, the method of correlation energy extrapolation by intrinsic scaling. Benchmark calculations of the new method are performed on diatomic fluorine and ozone. Finally, the method consistently achieves agreement with CI calculations to within a few mhartree and often achieves agreement to within ~1 millihartree or less, while requiring significantly less computational resources.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

A Co-Design Study Of Fusion Whole Device Modeling Using Code Coupling

Complex workflows consisting of multiple simulation and analysis codes running concurrently through in-memory coupling is becoming popular due to inherent advantages in online management of large-scale data, resilience, and the code development process. However, orchestrating such a multi-application workflow to efficiently utilize resources on a heterogeneous architecture is challenging. In this paper, we present our results with running the Fusion Whole Device Modeling benchmark workflow on Summit, a pre-exascale supercomputer at Oak Ridge National Laboratory. We explore various resource distribution and process placement mechanisms, including sharing compute nodes between processes from separate applications. We show that fine-grained process placement can have a significant impact towards efficient utilization of the compute power of a node on Summit, and conclude that sophisticated tools for performing co-design studies of multi-application workflows can play an important role towards efficient orchestration of such workflows.

Choi, Jong Youl↗

Embedded Volttron Specification - Benchmarking Small Footprint Compute Devices for Volttron

An embedded system is a small footprint computing unit that typically serves a specific purpose closely associated with measurements and control of hardware devices. These units are designed for reasonable durability and operations in a wide range of operating conditions. Some embedded systems support real-time operations and can demonstrate high levels of reliability. Many have failsafe mechanisms built to handle graceful shutdown of the device in exception conditions. The available memory, processing power, and network connectivity of these devices are limited due to the nature of their specific-purpose design and intended application. Industry practice is to carefully design the software for the available hardware capability to suit desired deployment needs. Volttron is an open source agent development and deployment platform designed to enable researchers to interact with devices and appliances without having to write drivers themselves. Hosting Volttron on small footprint embeddable devices enables its demonstration for embedded use. This report details the steps required and the experience in setting up and running Volttron applications on three small footprint devices: the Intel Next Unit of Computing (NUC), the Raspberry Pi 2, and the BeagleBone Black. In addition, the report also details preliminary investigation of the execution performance of Volttron on these devices.

97 MATHEMATICS AND COMPUTING↗

Parallel Agent-Based Simulations on Clusters of GPUs and Multi-Core Processors

An effective latency-hiding mechanism is presented in the parallelization of agent-based model simulations (ABMS) with millions of agents. The mechanism is designed to accommodate the hierarchical organization as well as heterogeneity of current state-of-the-art parallel computing platforms. We use it to explore the computation vs. communication trade-off continuum available with the deep computational and memory hierarchies of extant platforms and present a novel analytical model of the tradeoff. We describe our implementation and report preliminary performance results on two distinct parallel platforms suitable for ABMS: CUDA threads on multiple, networked graphical processing units (GPUs), and pthreads on multi-core processors. Message Passing Interface (MPI) is used for inter-GPU as well as inter-socket communication on a cluster of multiple GPUs and multi-core processors. Results indicate the benefits of our latency-hiding scheme, delivering as much as over 100-fold improvement in runtime for certain benchmark ABMS application scenarios with several million agents. This speed improvement is obtained on our system that is already two to three orders of magnitude faster on one GPU than an equivalent CPU-based execution in a popular simulator in Java. Thus, the overall execution of our current work is over four orders of magnitude faster when executed on multiple GPUs.

97 MATHEMATICS AND COMPUTING↗

EqualWrites: Reducing Intra-set Write Variations for Enhancing Lifetime of Non-volatile Caches

Driven by the trends of increasing core-count and bandwidth-wall problem, the size of last level caches (LLCs) has greatly increased and hence, the researchers have explored non-volatile memories (NVMs) which provide high density and consume low-leakage power. Since NVMs have low write-endurance and the existing cache management policies are write variation-unaware, effective wear-leveling techniques are required for achieving reasonable cache lifetimes using NVMs. We present EqualWrites, a technique for mitigating intra-set write variation. In this paper, our technique works by recording the number of writes on a block and changing the cache-block location of a hot data-item to redirect the future writes to a cold block to achieve wear-leveling. Simulation experiments have been performed using an x86-64 simulator and benchmarks from SPEC06 and HPC (high-performance computing) field. The results show that for single, dual and quad-core system configurations, EqualWrites improves cache lifetime by 6.31X, 8.74X and 10.54X, respectively. In addition, its implementation overhead is very small and it provides larger improvement in lifetime than three other intra-set wear-leveling techniques and a cache replacement policy.

97 MATHEMATICS AND COMPUTING↗

Si/SiGe QuBus for single electron information-processing devices with memory and micron-scale connectivity function

The connectivity within single carrier information-processing devices requires transport and storage of single charge quanta. Single electrons have been adiabatically transported while confined to a moving quantum dot in short, all-electrical Si/SiGe shuttle device, called quantum bus (QuBus). Here we show a QuBus spanning a length of 10 μm and operated by only six simply-tunable voltage pulses. We introduce a characterization method, called shuttle-tomography, to benchmark the potential imperfections and local shuttle-fidelity of the QuBus. The fidelity of the single-electron shuttle across the full device and back (a total distance of 19 μm) is (99.7 ± 0.3) %. Using the QuBus, we position and detect up to 34 electrons and initialize a register of 34 quantum dots with arbitrarily chosen patterns of zero and single-electrons. The simple operation signals, compatibility with industry fabrication and low spin-environment-interaction in 28 Si/SiGe, promises long-range spin-conserving transport of spin qubits for quantum connectivity in quantum computing architectures.

97 MATHEMATICS AND COMPUTING↗

HPC Node Performance and Energy Modeling with the Co-Location of Applications

Multicore processors have become an integral part of modern large-scale and high-performance parallel and distributed computing systems. This advance became necessary as the demand for increased system performance has exceeded the limits that single core processors can provide. Unfortunately, applications co-located on multicore processors can suffer from decreased performance and increased dynamic energy use as a result of access to shared resources, such as memory. Consequently, it is increasingly important to characterize the performance of applications that execute on these architectures. This work investigates some of the disadvantages of co-location, and presents a methodology for building models capable of utilizing varying amounts of information about a target application and its co-located applications to make predictions about the target application’s execution time and the system’s energy use under arbitrary co-locations of a wide range of application types. The proposed methodology is validated on three different server class Intel Xeon multicore processors using eleven applications from two scientific benchmark suites. The model’s utility for scheduling is also demonstrated in a simulated large-scale high-performance computing environment through the creation of a co-location aware scheduling heuristic. This heuristic demonstrates that scheduling using information generated with the proposed modeling methodology is capable of making significant improvements over a scheduling heuristic that is naive to co-location interference.

performance modeling, energy modeling, resource ma↗

Diagnosis and prognosis of Alzheimer's disease using brain morphometry and white matter connectomes

Accurate, reliable prediction of risk for Alzheimer's disease (AD) is essential for early, disease-modifying therapeutics. Multimodal MRI, such as structural and diffusion MRI, is likely to contain complementary information of neurodegenerative processes in AD. Here we tested the utility of the multimodal MRI (T1-weighted structure and diffusion MRI), combined with high-throughput brain phenotyping—morphometry and structural connectomics—and machine learning, as a diagnostic tool for AD. We used, firstly, a clinical cohort at a dementia clinic (National Health Insurance Service-Ilsan Hospital [NHIS-IH]; N = 211; 110 AD, 64 mild cognitive impairment [MCI], and 37 cognitively normal with subjective memory complaints [SMC]) to test the diagnostic models; and, secondly, Alzheimer's Disease Neuroimaging Initiative (ADNI)-2 to test the generalizability. Our machine learning models trained on the morphometric and connectome estimates (number of features = 34,646) showed optimal classification accuracy (AD/SMC: 97% accuracy, MCI/SMC: 83% accuracy; AD/MCI: 97% accuracy) in NHIS-IH cohort, outperforming a benchmark model (FLAIR-based white matter hyperintensity volumes). In ADNI-2 data, the combined connectome and morphometry model showed similar or superior accuracies (AD/HC: 96%; MCI/HC: 70%; AD/MCI: 75% accuracy) compared with the CSF biomarker model (t-tau, p-tau, and Amyloid β, and ratios). In predicting MCI to AD progression in a smaller cohort of ADNI-2 (n = 60), the morphometry model showed similar performance with 69% accuracy compared with CSF biomarker model with 70% accuracy. Our comparisons of the classifiers trained on structural MRI, diffusion MRI, FLAIR, and CSF biomarkers showed the promising utility of the white matter structural connectomes in classifying AD and MCI in addition to the widely used structural MRI-based morphometry, when combined with machine learning.

60 APPLIED LIFE SCIENCES↗

BCSR on GPU: A Way Forward Extreme-scale Graph Processing on Accelerator-enabled Frontier Supercomputer

Handling large graphs in a distributed environment requires effective partitioning across processors and efficient management of local partitions. In 2D partitioning, local graphs often become too sparse, making memory-efficient data structures crucial. Using the Compressed Sparse Row (CSR) format wastes space, especially for > 83% of vertices with empty edges for the sparse graphs. This study explores bit-CSR (BCSR), a modified CSR representation, on GPUs to reduce memory usage in graph computations. We achieved 16.67% memory savings on a sparse rmat dataset with 268 million vertices and 357 million edges, without performance degradation, supported by both theoretical and experimental storage savings of 33%. However, we observed a 1.7× slowdown in degree lookup times due to bitwise operations on AMD CPUs. This analysis highlights the potential of BCSR on GPUs for improving Graph500 benchmark performance on GPU-accelerated systems, such as the Frontier supercomputer.

Sattar, Naw Safrin↗

GPU-acceleration of the ELPA2 distributed eigensolver for dense symmetric and hermitian eigenproblems

The solution of eigenproblems is often a key computational bottleneck that limits the tractable system size of numerical algorithms, among them electronic structure theory in chemistry and in condensed matter physics. Large eigenproblems can easily exceed the capacity of a single compute node, thus must be solved on distributed-memory parallel computers. We here present GPU-oriented optimizations of the ELPA two-stage tridiagonalization eigensolver (ELPA2). On top of cuBLAS-based GPU offloading, we add a CUDA kernel to speed up the back-transformation of eigenvectors, which can be the computationally most expensive part of the two-stage tridiagonalization algorithm. Furthermore, we benchmark the performance of this GPU-accelerated eigensolver on two hybrid CPU–GPU architectures, namely a compute cluster based on Intel Xeon Gold CPUs and NVIDIA Volta GPUs, and the Summit supercomputer based on IBM POWER9 CPUs and NVIDIA Volta GPUs. Consistent with previous benchmarks on CPU-only architectures, the GPU-accelerated two-stage solver exhibits a parallel performance superior to the one-stage counterpart. Finally, we demonstrate the performance of the GPU-accelerated eigensolver developed in this work for routine semi-local KS-DFT calculations comprising thousands of atoms.

97 MATHEMATICS AND COMPUTING↗

A mass-conserving lattice Boltzmann method with dynamic grid refinement for immiscible two-phase flows

A mass-conserving lattice Boltzmann method (LBM) for multiphase flows is presented in this paper. The proposed LBM improves a previous model (Lee and Liu, 2010 [21]) in terms of mass conservation, speed-up, and efficiency, and also extends its capabilities for implementation on non-uniform grids. The presented model consists of a phase-field lattice Boltzmann equation (LBE) for tracking the interface between different fluids and a pressure-evolution LBM for recovering the hydrodynamic properties. In addition to the mass conservation property and the simplicity of the algorithm, the advantages of the current phase-field LBE are that it is an order of magnitude faster than the previous interface tracking LBE proposed by Lee and Liu (2010) [21] and it requires less memory resources for data storage. Meanwhile, the pressure-evolution LBM is equipped with a multi-relaxation-time (MRT) collision operator to facilitate attainability of small relaxation rates thereby allowing simulation of multiphase flows at higher Reynolds numbers. Additionally, we reformulate the presented MRT-LBM on nonuniform grids within an adaptive mesh refinement (AMR) framework. Various benchmark studies such as a rising bubble and a falling drop under buoyancy, droplet splashing on a wet surface, and droplet coalescence onto a fluid interface are conducted to examine the accuracy and versatility of the proposed AMR-LBM. The proposed model is further validated by comparing the results with other LB models on uniform grids. A factor of about 20 in savings of computational resources is achieved by using the proposed AMR-LBM. As a more demanding application, the Kelvin–Helmholtz instability (KHI) of a shear-layer flow is investigated for both density-matched and density-stratified binary fluids. The KHI results of the density-matched fluids are shown to be in good agreement with the benchmark AMR results based on the sharp-interface approach. When a density contrast between the two fluids exists, a typical chaotic structure in the flow field is observed at a Reynolds number of 10000, which indicates that the proposed model is a promising tool for direct numerical simulation of two-phase flows.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

Second-Order Møller–Plesset Perturbation Theory in the Condensed Phase: An Efficient and Massively Parallel Gaussian and Plane Waves Approach

A novel algorithm, based on a hybrid Gaussian and plane waves (GPW) approach, is developed for the canonical second-order Møller–Plesset perturbation energy (MP 2 ) of finite and extended systems. The key aspect of the method is that the electron repulsion integrals (ia|λσ) are computed by direct integration between the products of Gaussian basis functions λσ and the electrostatic potential arising from a given occupied-virtual pair density ia. The electrostatic potential is obtained in a plane waves basis set after solving the Poisson equation in Fourier space. In particular, for condensed phase systems, this scheme is highly efficient. Furthermore, our implementation has low memory requirements and displays excellent parallel scalability up to 100 000 processes. In this way, canonical MP 2 calculations for condensed phase systems containing hundreds of atoms or more than 5000 basis functions can be performed within minutes, while systems up to 1000 atoms and 10 000 basis functions remain feasible. Solid LiH has been employed as a benchmark to study basis set and system size convergence. Lattice constants and cohesive energies of various molecular crystals have been studied with MP 2 and double-hybrid functionals.

Chemistry↗

Multi-Source Data Aggregation and Real-Time Anomaly Classification and Localization in Power Distribution Systems

This paper proposes a real-time anomaly location and classification framework for power distribution systems to simultaneously determine the type of anomaly (i.e., short-circuit fault, cyber attack, DER switching) and its location. The proposed framework employs the data aggregation module to collect the measurement data from multiple field devices operating at different sampling rates, such as protection relays and D-PMUs. The output of the data aggregation is then fed into a multi-task learning-based long-based short-term memory (MTL-LSTM) to classify the type of anomaly and the location in two separate tasks. The proposed MTL-LSTM approach can be utilized in real-time operation in order to distinguish between normal and several anomalous operations and locate the anomaly. The proposed framework is tested on a modified IEEE 33-bus test feeder benchmark that integrates solar generation and energy storage. Furthermore, the results show that the proposed framework can locate and classify anomalies for several operation conditions with more than 96% accuracy. Further experiments highlight the impact of aggregating multiple sources of data on the performance of the proposed model.

24 POWER TRANSMISSION AND DISTRIBUTION↗