Search NASA⌕ Search

SEARCH · Search NASA

Results for “Computational Load Balancing”

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 703 records · Page 39

Acceleration of Semiempirical QM/MM Methods through Message Passage Interface (MPI), Hybrid MPI/Open Multiprocessing, and Self-Consistent Field Accelerator Implementations

The strategy and implementation of scalable and efficient semiempirical (SE) QM/MM methods in CHARMM are described. The serial version of the code was first profiled to identify routines that required parallelization. Afterward, the code was parallelized and accelerated with three approaches. The first approach was the parallelization of the entire QM/MM routines, including the Fock matrix diagonalization routines, using the CHARMM message passage interface (MPI) machinery. In the second approach, two different self-consistent field (SCF) energy convergence accelerators were implemented using density and Fock matrices as targets for their extrapolations in the SCF procedure. In the third approach, the entire QM/MM and MM energy routines were accelerated by implementing the hybrid MPI/open multiprocessing (OpenMP) model in which both the task- and loop-level parallelization strategies were adopted to balance loads between different OpenMP threads. The present implementation was tested on two solvated enzyme systems (including <100 QM atoms) and an S N 2 symmetric reaction in water. The MPI version exceeded existing SE QM methods in CHARMM, which include the SCC-DFTB and SQUANTUM methods, by at least 4-fold. The use of SCF convergence accelerators further accelerated the code by ~12–35% depending on the size of the QM region and the number of CPU cores used. Although the MPI version displayed good scalability, the performance was diminished for large numbers of MPI processes due to the overhead associated with MPI communications between nodes. This issue was partially overcome by the hybrid MPI/OpenMP approach which displayed a better scalability for a larger number of CPU cores (up to 64 CPUs in the tested systems).

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

Model-based Assessment for Balancing Privacy Requirements and Operational Capabilities

The smart grid changes the way energy is produced and distributed. In addition both, energy and information is exchanged bidirectionally among participating parties. Therefore heterogeneous systems have to cooperate effectively in order to achieve a common high-level use case, such as smart metering for billing or demand response for load curtailment. Furthermore, a substantial amount of personal data is often needed for achieving that goal. Capturing and processing personal data in the smart grid increases customer concerns about privacy and in addition, certain statutory and operational requirements regarding privacy aware data processing and storage have to be met. An increase of privacy constraints, however, often limits the operational capabilities of the system. In this paper, we present an approach that automates the process of finding an optimal balance between privacy requirements and operational requirements in a smart grid use case and application scenario. This is achieved by formally describing use cases in an abstract model and by finding an algorithm that determines the optimum balance by forward mapping privacy and operational impacts. For this optimal balancing algorithm both, a numeric approximation and – if feasible – an analytic assessment are presented and investigated. The system is evaluated by applying the tool to a real-world use case from the University of Southern California (USC) microgrid.

29 ENERGY PLANNING, POLICY, AND ECONOMY↗

Multi-cycle reload analysis of a long cycle gas-cooled fast modular reactor

There is currently significant interest in deploying HALEU-fueled fast reactors, including the General Atomics (GA) Fast Modular Reactor (FMR). Such reactors can achieve very long fuel cycles, but with multi-batch loading will take decades to reach equilibrium. This motivates design and analysis of both the initial core and multi-cycle reload, which is typically performed using fast-running, deterministic fast reactor codes such as the Argonne Reactor Computation (ARC) codes. In this paper, multicycle reload of the GA FMR is analyzed using the ARC codes. The GA FMR utilizes 19.75 % enriched fuel in a 16 year cycle with a three-batch strategy, with twice-burned fuel placed on the core periphery. The GA FMR has a softened neutron spectrum due to reflecting elements in the core, so the neutronic solution is first benchmarked against the OpenMC Monte Carlo code. Discrepancy on k eff is 400–600 pcm, likely due to the softened neutron spectrum, heterogeneous fuel assembly design and central reflector. However, the rms discrepancy on the assembly power distribution is only 0.6 %, despite the presence of the central reflector. A reload strategy is devised for the first three cycles of such a reactor, ultimately spanning the first 45–48 years of its operation. The fresh core uses 19.75 %, 19.25 % and 16.75 % enriched fuel in place of fresh, once-burned and twice-burned and is then subsequently refueled with only 19.75 % enriched fuel. The cycle length is varied over 3 cycles of operation to balance fuel utilization and reactor availability, specifically with use of an extended 18-year Cycle 1, followed by a shortened 11-year Cycle 2. Cycle 3 is close to the target 16-year length. Finally, placing twice burned assemblies next to the GA FMR central reflector can reduce power peaking by 3 %, at the expense of slightly reducing the cycle length.

22 GENERAL STUDIES OF NUCLEAR REACTORS↗

Automatic and Transparent Resource Contention Mitigation for Improving Large-Scale Parallel File System Performance

Proportional to the scale increases in HPC systems, many scientific applications are becoming increasingly data intensive, and parallel I/O has become one of the dominant factors impacting the large-scale HPC application performance. On a typical large-scale HPC system, we have observed that the lack of a global workload coordination coupled with the shared nature of storage systems cause load imbalance and resource contention over the end-to-end I/O paths resulting in severe performance degradation. I/O load imbalance on HPC systems is generally a self-inflicted wound and mostly occurs between the I/O paths and resources consumed by each individual job. In this paper, we introduce TAPP-IO, a dynamic, shared load balancing framework for mitigating resource contention. TAPP-IO extends our previous work and solves two major limitations: First, it transparently intercepts file creation calls during runtime to balance the workload over all available storage targets. The usage of TAPP-IO requires no application source code modifications and is independent from any I/O middleware. The framework can be applied to almost any HPC platform and is suitable for systems that lack a centralized file system resource manager. Second, the framework proposes a new placement strategy to support not only file-per-process I/O, but also single shared file I/O. This opens the door to a new class of scientific applications that can leverage the placement library for improved performance. We demonstrate the effectiveness of our integration on the Titan system at the Oak Ridge National Laboratory. Our experiments with a synthetic benchmark and real-world HPC workload show that, even in a noisy production environment, TAPP-IO can improve large-scale application performance significantly.

Neuwirth, Sarah↗

Influence of Strain-Gage Balance Characteristics on the Precision Error of the Drag Coefficient

Improvements and results of a new method are presented that computes a pre-test estimate of the precision error of the drag coefficient of a wind tunnel model. The error estimate is defined as the part of the drag coefficient's precision error that is primarily associated with the precision error of the angle of attack measurement and physical characteristics of the chosen strain-gage balance. The method indirectly describes the precision error of the angle of attack measurement by using an assumed balance gage output variation of one microV/V. The physical characteristics of the balance, on the other hand, are described by partial derivatives of the axial and normal forces with respect to the strain-gage outputs. These derivatives can directly be obtained from the data reduction matrix of the balance. The precision error estimate itself is calculated by applying a simple explicit equation that uses the model reference area, the dynamic pressure, the angle of attack, the coefficients of the linear terms of the data reduction matrix, and the electrical output variation of one microvolt per volt as input. Precision errors at constant angle of attack may be visualized as contour plots by plotting them, for example, versus the Mach number and the total pressure. Characteristics of NASA's MC60E balance are used in combination with the reference area of a generic wind tunnel model in order to demonstrate that error estimates are independent of both the balance load format and the units chosen for the description of balance loads, model reference area, and the dynamic pressure. Finally, experimental data from a wind tunnel test of the Ames Check Standard Model in the NASA Ames 11-foot Transonic Wind Tunnel illustrates the application of the method to real-world test data.

precision error↗

Receptor tyrosine kinase inhibition causes simultaneous bone loss and excess bone formation within growing bone in rats

During postnatal skeletal growth, adaptation to mechanical loading leads to cellular activities at the growth plate. It has recently become evident that bone forming and bone resorbing cells are affected by the receptor tyrosine kinase (RTK) inhibitor imatinib mesylate (STI571, Gleevec (registered)) . Imatinib targets PDGF, ABL-related gene, c-Abl, c-Kit and c-Fms receptors, many of which have multiple functions in the bone microenvironment. We therefore studied the effects of imatinib in growing bone. Young rats were exposed to imatinib (150 mg/kg on postnatal days 5-7, or 100 mg/kg on postnatal days 5-13), and the effects of RTK inhibition on bone physiology were studied after 8 and 70 days (3-day treatment), or after 14 days (9-day treatment). X-ray imaging, computer tomography, histomorphometry, RNA analysis and immunohistochemistry were used to evaluate bone modeling and remodeling in vivo. Imatinib treatment eliminated osteoclasts from the metaphyseal osteochondral junction at 8 and 14 days. This led to a resorption arrest at the growth plate, but also increased bone apposition by osteoblasts, thus resulting in local osteopetrosis at the osteochondral junction. The impaired bone remodelation observed on day 8 remained significant until adulthood. Within the same bone, increased osteoclast activity, leading to bone loss, was observed at distal bone trabeculae on days 8 and 14. Peripheral quantitative computer tomography (pQCT) and micro-CT analysis confirmed that, at the osteochondral junction, imatinib shifted the balance from bone resorption towards bone formation, thereby altering bone modeling. At distal trabecular bone, in turn, the balance was turned towards bone resorption, leading to bone loss. - Research Highlights: > 3-Day imatinib treatment. > Causes growth plate anomalies in young rats. > Causes biomechanical changes and significant bone loss at distal trabecular bone. > Results in loss of osteoclasts at osteochondral junction.

60 APPLIED LIFE SCIENCES↗

Sensitivity analysis for the dynamic response of thermoviscoplastic shells of revolution

A computational procedure is presented for evaluating the sensitivity coefficients of the dynamic axisymmetric, fully-coupled, thermoviscoplastic response of shells of revolution. The analytical formulation is based on Reissner's large deformation shell theory with the effects of large-strain, transverse shear deformation, rotatory inertia and moments turning around the normal to the middle surface included. The material model is chosen to be viscoplasticity with strain hardening and thermal hardening, and an associated flow rule is used with a von Mises effective stress. A mixed formulation is used for the shell equations with the fundamental unknowns consisting of six stress resultants, three generalized displacements and three velocity components. The energy-balance equation is solved using a Galerkin procedure, with the temperature as the fundamental unknown. Spatial discretization is performed in one dimension (meridional direction) for the momentum and constitutive equations of the shell, and in two dimensions (meridional and thickness directions) for the energy-balance equation. The temporal integration is performed by using an explicit central difference scheme (leap-frog method) for the momentum equation; a predictor-corrector version of the trapezoidal rule is used for the energy-balance equation; and an explicit scheme consistent with the central difference method is used to integrate the constitutive equations. The sensitivity coefficients are evaluated by using a direct differentiation approach. Numerical results are presented for a spherical cap subjected to step loading.

Kulkarni, Makarand↗

MEITNER Resource Team Modeling and Simulation Support to Holos-Quad Reactor Development (Final CRADA Report)

The main objective of this CRADA is to provide Argonne National Laboratory (ANL)’s modeling and simulation capabilities via ARPA-E's MEITNER (Modeling-Enhanced Innovations Trailblazing Nuclear Energy Reinvigoration) Resource Team (RT) arrangement to support the demonstration of the viability of HolosGen’s Holos-Quad reactor design. The Holos-Quad reactor design is an advanced reactor concept that incorporates many new design features, such as the neutron-coupled Subcritical Power Modules (SPMs) in its core design, elimination of balance of plant (BOP) by direct integration of a helium Brayton cycle power conversion system with each SPM, among many other innovative features. The demonstration of the viability of such an innovative reactor design warrants iterations of modeling and simulation and testing. The purpose of this project is to utilize ANL’s modeling and simulation capabilities in nuclear reactor analysis and power conversion system analysis to inform HolosGen and the Design Team (DT) in the design and optimization of the Holos-Quad concept. The major work scopes of this project include: to investigate conceptual designs and materials for radiation shielding to protect personnel and internal components such as turbomachinery; Assess the the helium Brayton cycle power conversion system performance in both nominal and load following conditions; Investigate power conversion components and overall system performance; Identify control strategies to enable load following; Perform simulations of HolosGen’s subscale simulator and using available test data for code validation/benchmark purpose; Perform core thermal-hydraulics, safety analysis, and structural analysis.

21 SPECIFIC NUCLEAR REACTORS AND ASSOCIATED PLANTS↗

The Data Transfer Kit: A geometric rendezvous-based tool for multiphysics data transfer

The Data Transfer Kit (DTK) is a software library designed to provide parallel data transfer services for arbitrary physics components based on the concept of geometric rendezvous. The rendezvous algorithm provides a means to geometrically correlate two geometric domains that may be arbitrarily decomposed in a parallel simulation. By repartitioning both domains such that they have the same geometric domain on each parallel process, efficient and load balanced search operations and data transfer can be performed at a desirable algorithmic time complexity with low communication overhead relative to other types of mapping algorithms. With the increased development efforts in multiphysics simulation and other multiple mesh and geometry problems, generating parallel topology maps for transferring fields and other data between geometric domains is a common operation. The algorithms used to generate parallel topology maps based on the concept of geometric rendezvous as implemented in DTK are described with an example using a conjugate heat transfer calculation and thermal coupling with a neutronics code. In addition, we provide the results of initial scaling studies performed on the Jaguar Cray XK6 system at Oak Ridge National Laboratory for a worse-case-scenario problem in terms of algorithmic complexity that shows good scaling on 0(1 x 104) cores for topology map generation and excellent scaling on 0(1 x 105) cores for the data transfer operation with meshes of O(1 x 109) elements. (authors)

97 MATHEMATICS AND COMPUTING↗

Modeling Universal Globally Adaptive Load-Balanced Routing

Universal globally adaptive load-balanced (UGAL) routing has been proposed for various interconnection networks and has been deployed in a number of current-generation supercomputers. Although UGAL-based schemes have been extensively studied, most existing results are based on either simulation or measurement. Without a theoretical understanding of UGAL, multiple questions remain: For which traffic patterns is UGAL most suited? Furthermore, what determines the performance of the UGAL-based scheme on a particular network configuration? Here, we develop a set of throughput models for UGALbased on linear programming. We show that the throughput models are valid across the torus, Dragonfly, and Slim Fly network topologies. Finally, we identify a robust model that can accurately and efficiently predict UGAL throughput for a set of representative traffic patterns across different topologies. Our models not only provide a mechanism to predict UGAL performance on large-scale interconnection networks but also reveal the inner working of UGAL and further our understanding of this type of routing.

97 MATHEMATICS AND COMPUTING↗

iHarmonizer: improving the disk efficiency of I/O-intensive multithreaded codes

Challenged by serious power and thermal constraints, and limited by available instruction-level parallelism, processor designs have evolved to multi-core architectures. These architectures, many augmented with native simultaneous multithreading, are driving software developers to use multithreaded programs to exploit thread-level parallelism. While multithreading is well-known to introduce concerns of data dependency and CPU load balance, less known is that the uncertainty of relative progress of thread execution can cause patterns of I/O requests, issued by different threads, to be effectively random and so significantly degrade hard-disk efficiency. This effect can severely offset the performance gains from parallel execution, especially for I/O-intensive programs. Retaining the benefits of multithreading while not reducing I/O efficiency is an urgent and challenging problem. We propose a user-level scheme, iHarmonizer, to streamline the servicing of I/O requests from multiple threads in OpenMP programs. Specifically, we use the compiler to insert code into OpenMP programs so that data usage can be transmitted at run time to a supporting run-time library; this library in turn prefetches data in a disk-friendly way and coordinates threads execution according to the availability of their requested data. Transparently to the programmer, iHarmonizer makes a multithreaded program I/O efficient while maintaining the benefits of parallelism. Our experiments show that iHarmonizer can significantly speed up the execution of a representative set of I/O-intensive scientific benchmarks.

97 MATHEMATICS AND COMPUTING↗

Energy-efficient cooperative resource allocation and task scheduling for Internet of Things environments

Offloading Internet of Things (IoT) tasks to the cloud for further processing might not always lead to an optimal execution time, particularly in situations such as resource contention, under-provisioning, over-provisioning, and fragmentation. In addition, dynamically optimizing the number of Virtual Machines (VMs) for resource scheduling in order to meet application requirements remains a major research challenge. Further, existing resource scheduling algorithms focus primarily on minimizing operational costs while maximizing resource sharing and utilization. Considering energy utilization as part of the resource allocation and scheduling process as an optimization objective for maintaining load balancing has often been neglected. To address these challenges and more, we propose a cooperative energy-aware resource allocation and scheduling strategy based on a Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) multi-criteria decision-making method. Here we used the Grid Workloads Archive dataset to evaluate our proposed approach named TOPREAL. Experimental results with respect to the allocation of VM resources when considering processing a large segment of tasks indicate that TOPREAL outperforms existing algorithms in terms of energy savings, with an average improvement of 40.25%, while maintaining an average improvement of 16.21% when it comes to execution time. Results also demonstrate that our method can save an average of 78.06 processing hours and 63,215kJ of energy when compared to existing scheduling algorithms. These results demonstrate the effectiveness of our proposed model and the viability of using multi-criteria decision-making techniques such as TOPSIS to solve the resource allocation and scheduling problem in edge environments.

32 ENERGY CONSERVATION, CONSUMPTION, AND UTILIZATI↗

Probing the Energetics of Molecule–Material Interactions at Interfaces and in Nanopores

During the past decades, advances in interfacial chemistries at the molecular level are shaping our world by playing crucial roles in balancing global scale energy crisis and critical environmental concerns. However, systematic investigations into the binding energies, site distribution and their correlation with the molecular-level surface assemblages and structures at interfaces and in nanopores are rarely documented. In this review, we summarize a set of systematic calorimetric studies on surface energetics we performed during the last decade. These studies demonstrate how thermochemistry can reveal crucial energetic insights into a series of molecule – material interactions relevant to a number of applications, including carbon capture and sequestration, energy production, sustainable chemical processing, catalysis, and nanogeoscience. Calorimetric methodologies developed and applied include direct gas adsorption calorimetry, nearroom temperature solvent immersion/solution calorimetry and high temperature oxide melt solution calorimetry. Using these highly unique techniques, we reveal the thermodynamic complexity of carbon dioxide capture on metal – organic framework (MOF) sorbents with built-in and grafted nucleophilic functional groups (-OH and -NH2). These studies suggest that carbon dioxide adsorption on functionalized MOFs is a complex process involving multiple thermodynamic factors, as reflected by changes in surface phase and structure, chemical bonding and degree of disorder with varying temperature and gas loading. The fundamental insights obtained may help optimize the design, synthesis and application of MOF-based carbon dioxide sorbents for carbon capture and sequestration. In parallel, we also explore the energetics of interaction and competition between small molecules (water, carbon dioxide, methane, simple and complex organics) and inorganic materials (calcite, silica, zirconia, zeolites, mesoporous frameworks, alumina, and uranium), at interfaces and in nanopores. Combined with spectroscopic, diffraction, electron microscopic and computational techniques, the energetics of gas/liquid – solid interactions can be correlated with specific bonds, molecular configurations and nanostructures. Although the energetics evolves continuously from weak association to strong bonding to classical capping, distinct regions of rapidly changing stepwise energetics often separate the different regimes. These phenomena are closely related to the properties of inorganic material surfaces (hydrophobicity and acidity/basicity), the framework architectures, and the chemical nature of adsorbate molecules. These direct thermodynamic insights reinforce our understanding of complex small molecule – inorganic material interactions important to multiple disciplines of chemical engineering, materials science, nanogeoscience and environmental technology, including heterogeneous catalysis, molecular separation, material design and synthesis, biomineralization, contaminant and nutrient transport, carbonate formation, and water – organic competitions on material/mineral surfaces.

37 INORGANIC, ORGANIC, PHYSICAL, AND ANALYTICAL CH↗

HydraGNN_Predictive_GFM_2024 - Ensemble of predictive graph foundation models for ground state atomistic materials modeling

We provide the ensemble of fifteen pre-trained graph foundation models (GFMs) for atomistic materials modeling applications. Each one of the fifteen GFMs has been trained on five open-source datasets that (once aggregated) amount to over 154 million atomistic structures, which cover over two-thirds of the natural elements of the periodic table and that comprises a broad set of organic and inorganic compounds. This vast set of atomistic structures comprises ground state configurations that are dynamically stable (i.e., equilibrated structures with atomic forces approximately close to zero values) as well as dynamically unstable structures (i.e., non-equilibrium structures with non-negligible non-zero values of atomic forces). The ensemble of datasets aggregated does NOT include excited states. The datasets have been curated to remove atomistic structures with spectral norm of the force tensor above 100 eV/angstrom. Moreover, a linear term of the energy was computed for each dataset using a linear regression model that uses the chemical concentration of each natural element as regressor. The linear term predicted by the linear regression model has been subtracted from each original energy value to perform a re-alignment of the energy values across different electronic structures approximation theories performed to generate the diverse multi-source, multi-fidelity datasets. The folder "ADIOS_files" contains the set of pre-processed datasets in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used for the development and training of GFMs in this work. The "ADIOS_files" directory contains 6 sub-directories named as follows: - ANI1x-v3.bp - MPTrj-v3.bp - OC2020-20M-v3.bp - OC2020-v3.bp - OC2022-v3.bp - qm7x-v3.bp Each sub-directory contains the pre-processed datasets converted in Adaptable I/O System (ADIOS) format (https://www.exascaleproject.org/research-project/adios/) that have been used to the development, training, and performance testing of the ensemble go predictive graph foundation models. Each GFM was developed using HydraGNN (https://github.com/ORNL/HydraGNN) as underlying graph neural network (GNN) architecture. The multi-task learning (MTL) capability of HydraGNN was used to simultaneously train the GFMs on labeled values for direct predictions of energy (a total system property of an atomistic structure that measures the chemical stability) and atomic forces (an atomic level property of an atomistic structure that measures the dynamical stability). The hyper parameters of the GFM have been tuned using scalable hyperparameter optimization (HPO) algorithms implemented in the software DeepHyper (https://github.com/deephyper/deephyper). The pre-training of each HPO trial was performed using distributed data parallelism (DDP) to scale the training across 128 compute nodes of the exascale OLCF supercomputer Frontier. Each HPO trial was trained only for 10 epochs and an early stopping was performed to avoid wasting significant computational resources on GNN architectures that were clearly underperforming. For each HPO trial, the 'omnistat' tool developed by (AMD Research - Advanced Micro Device) was used to measure the total energy consumption in kWh. The ensemble of GFMs was obtained by selecting the fifteen best performing HPO trials. Four models have been selected for their clear advantage in accuracy, and these are the GFMs with IDs 229, 156, 147, 260. Additional eleven models have been selected based on judicious balance between accuracy and energy consumption needed for training, and these are the GFMs with IDs 165, 78, 137, 1, 175, 171, 181, 67, 179, 167, 351. Each selected GFM of the ensemble was continued to cumulate a total of at most 30 epochs. In some cases, the total number of epochs actually performed was les than 30 due to two combined factors: (1) the size of the GFM (i.e., the number of model parameters to train) and (2) the total wall-clock time for which the computational resources could be allocated on OLCF-Frontier. The "Ensemble_of_models" directory contains 15 sub-directories named as follows: - gfm_0.229 - gfm_0.156 - gfm_0.147 - gfm_0.260 - gfm_0.165 - gfm_0.78 - gfm_0.137 - gfm_0.1 - gfm_0.175 - gfm_0.171 - gfm_0.181 - gfm_0.67 - gfm_0.179 - gfm_0.167 - gfm_0.351 Each one of these sub-directories refers to one of the fifteen HPO trials that have been selected to continue the pre-training with at most 30 epochs. With each sub-directory associated with a specific HPO trial, the following files can be found: - config.json: file for argument parsing to develop and train an HydraGNN architecture - gfm_0.ID_epoch_N.pk: file with model parameters for HPO ID trial after N epochs of training The ensemble of fifteen GFM architectures was used for (1) ensemble averaging to stabilize the predictions of energy and atomic forces after pre-training for post-processing analysis and (2) ensemble uncertainty quantification (UQ). The code used to develop, pre-train, and load the pre-trained models for post-processing analysis is available on the ORNL-GitHub at the following link: https://github.com/ORNL/HydraGNN/tree/Predictive_GFM_2024

36 MATERIALS SCIENCE↗

Design and implementation of scalable tape archiver

In order to reduce costs, computer manufacturers try to use commodity parts as much as possible. Mainframes using proprietary processors are being replaced by high performance RISC microprocessor-based workstations, which are further being replaced by the commodity microprocessor used in personal computers. Highly reliable disks for mainframes are also being replaced by disk arrays, which are complexes of disk drives. In this paper we try to clarify the feasibility of a large scale tertiary storage system composed of 8-mm tape archivers utilizing robotics. In the near future, the 8-mm tape archiver will be widely used and become a commodity part, since recent rapid growth of multimedia applications requires much larger storage than disk drives can provide. We designed a scalable tape archiver which connects as many 8-mm tape archivers (element archivers) as possible. In the scalable archiver, robotics can exchange a cassette tape between two adjacent element archivers mechanically. Thus, we can build a large scalable archiver inexpensively. In addition, a sophisticated migration mechanism distributes frequently accessed tapes (hot tapes) evenly among all of the element archivers, which improves the throughput considerably. Even with the failures of some tape drives, the system dynamically redistributes hot tapes to the other element archivers which have live tape drives. Several kinds of specially tailored huge archivers are on the market, however, the 8-mm tape scalable archiver could replace them. To maintain high performance in spite of high access locality when a large number of archivers are attached to the scalable archiver, it is necessary to scatter frequently accessed cassettes among the element archivers and to use the tape drives efficiently. For this purpose, we introduce two cassette migration algorithms, foreground migration and background migration. Background migration transfers cassettes between element archivers to redistribute frequently accessed cassettes, thus balancing the load of each archiver. Background migration occurs the robotics are idle. Both migration algorithms are based on access frequency and space utility of each element archiver. To normalize these parameters according to the number of drives in each element archiver, it is possible to maintain high performance even if some tape drives fail. We found that the foreground migration is efficient at reducing access response time. Beside the foreground migration, the background migration makes it possible to track the transition of spatial access locality quickly.

Nemoto, Toshihiro↗

Data-driven Fault Tolerance for Work Stealing Computations

Checkpoint-restart approaches to fault tolerance typically roll back all the processes to the previous checkpoint in the event of a failure. Work stealing is a promising technique to dynamically tolerate variations in the execution environment, including faults, system noise, and energy constraints. In this paper, we present fault tolerance mechanisms for task parallel computations, a popular computation idiom, employing work stealing. The computation is organized as a collection of tasks with data in a global address space. The completion of data operations, rather than the actual messages, is tracked to derive an idempotent data store. This information is used to accurately identify the tasks to be re-executed, therefore to recompute only the lost data, in the presence of random work stealing. We consider three recovery schemes that present distinct trade-offs -- lazy recovery with potentially increased re-execution cost, immediate collective recovery with associated synchronization overheads, and noncollective recovery enabled by additional communication. We employ distributed work stealing to dynamically rebalance the tasks on the live processes and evaluate the three schemes using candidate application benchmarks. We demonstrate that the overheads (space and time) of the fault tolerance mechanism are low, the cost incurred due to failures are small, and the overheads decrease with per-process work at scale.

fault tolerance, task parallelism, load balancing,↗

A scalable geometric multigrid solver for nonsymmetric elliptic systems with application to variable-density flows

A geometric multigrid algorithm is introduced for solving nonsymmetric linear systems resulting from the discretization of the variable density Navier–Stokes equations on nonuniform structured rectilinear grids and high-Reynolds number flows. The restriction operation is defined such that the resulting system on the coarser grids is symmetric, thereby allowing for the use of efficient smoother algorithms. To achieve an optimal rate of convergence, the sequence of interpolation and restriction operations are determined through a dynamic procedure. A parallel partitioning strategy is introduced to minimize communication while maintaining the load balance between all processors. To test the proposed algorithm, we consider two cases: 1) homogeneous isotropic turbulence discretized on uniform grids and 2) turbulent duct flow discretized on stretched grids. Testing the algorithm on systems with up to a billion unknowns shows that the cost varies linearly with the number of unknowns. This O (N) behavior confirms the robustness of the proposed multigrid method regarding ill-conditioning of large systems characteristic of multiscale high-Reynolds number turbulent flows. The robustness of our method to density variations is established by considering cases where density varies sharply in space by a factor of up to 104, showing its applicability to two-phase flow problems. Strong and weak scalability studies are carried out, employing up to 30,000 processors, to examine the parallel performance of our implementation. Excellent scalability of our solver is shown for a granularity as low as 104 to 105 unknowns per processor. At its tested peak throughput, it solves approximately 4 billion unknowns per second employing over 16,000 processors with a parallel efficiency higher than 50%.

Computer Science↗

Multi-Agent Hierarchical Deep Reinforcement Learning for HVAC Control With Flexible DERs

As electricity consumption in commercial and residential buildings continues to rise, reducing energy costs presents an increasing challenge. Heating, ventilating, and air-conditioning (HVAC) systems, which typically account for 40%-50% of a building's energy use, are prime targets for energy savings. Intelligent control of HVAC temperature through the exploitation of HVAC load flexibility brings significant potential to reduce energy consumption and electricity expenses. The nonlinear models of HVAC systems challenge traditional control methods, while the uncertainty introduced by HVAC load flexibility complicates distributed energy resource (DER) management using conventional optimal dispatch techniques. In response to these challenges, we propose a hierarchical multi-agent deep reinforcement learning (DRL) approach. The lower-level agents focus on balancing comfort and energy conservation, while the upper-level DRL agents optimize the use of DERs to reduce peak demand based on the control outcomes of the HVAC by the lower-level agents. Here, in the upper-level agents, we incorporate a multi-agent structure based on ensemble learning, which acts based on historical and current data without relying on precise load forecasting to address the delayed rewarding issue in DRL. This allows for the effective reduction of energy costs. The proposed method is tested using a real-world microgrid comprising 413 buildings in Southern California, and the results demonstrate that our approach can significantly reduce overall electricity bills while ensuring the comfort of consumers and residents.

24 POWER TRANSMISSION AND DISTRIBUTION↗