Search NASA⌕ Search

SEARCH · Search NASA

Results for “HPC system”

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 235 records · Page 13

Perspectives on AI Architectures and Co-design for Earth System Predictability

Recently, the U.S. Department of Energy (DOE), Office of Science, Biological and Environmental Research (BER), and Advanced Scientific Computing Research (ASCR) programs organized and held the Artificial Intelligence for Earth System Predictability (AI4ESP) workshop series. From this workshop, a critical conclusion that the DOE BER and ASCR community came to is the requirement to develop a new paradigm for Earth system predictability focused on enabling artificial intelligence (AI) across the field, laboratory, modeling, and analysis activities, called model experimentation (ModEx). BER’s ModEx is an iterative approach that enables process models to generate hypotheses. The developed hypotheses inform field and laboratory efforts to collect measurement and observation data, which are subsequently used to parameterize, drive, and test model (e.g., process based) predictions. A total of 17 technical sessions were held in this AI4ESP workshop series. This paper discusses the topic of the AI Architectures and Codesign session and associated outcomes. The AI Architectures and Codesign session included two invited talks, two plenary discussion panels, and three breakout rooms that covered specific topics, including 1) DOE high-performance computing (HPC) systems, 2) cloud HPC systems, and 3) edge computing and Internet of Things (IoT). We also provide forward-looking ideas and perspectives on potential research in this codesign area that can be achieved by synergies with the other 16 session topics. These ideas include topics such as 1) reimagining codesign, 2) data acquisition to distribution, 3) heterogeneous HPC solutions for integration of AI/ML and other data analytics like uncertainty quantification with Earth system modeling and simulation, and 4) AI-enabled sensor integration into Earth system measurements and observations. Such perspectives are a distinguishing aspect of this paper.

58 GEOSCIENCES↗

Quantifying the Impact of Advanced Web Platforms on High Performance Computing Usage

The deployment of Science Gateways for High Performance Computing (HPC) systems can alter long-accepted usage patterns on supercomputing systems in positive ways as an ever-increasing number of users migrate their workflows to HPC systems. Idaho National Laboratory (INL) has deployed two separate advanced web platforms, Open OnDemand and NICE DCV, for integration with HPC resources to improve web accessibility for HPC users. Researchers conducted a multi-year study on how HPC usage pat- terns changed in the presence of these platforms. This work reports the results of that study and quantifies the observed impacts, including adoption by visualization and Jupyter Notebook/Lab users, decreased job submission friction, rapid uptake of HPC by Windows users, and increased overall system utilization. The most significant impacts were observed from the deployment of Open OnDemand, and this work also identifies some best practices for Open OnDemand deployment for HPC datacenters.

97 MATHEMATICS AND COMPUTING↗

Scaling Deep Learning on GPU and Knights Landing clusters

The speed of deep neural networks training has become a big bottleneck of deep learning research and development. For example, training GoogleNet by ImageNet dataset on one Nvidia K20 GPU needs 21 days. To speed up the training process, the current deep learning systems heavily rely on the hardware accelerators. However, these accelerators have limited on-chip memory compared with CPUs. To handle large datasets, they need to fetch data from either CPU memory or remote processors. We use both self-hosted Intel Knights Landing (KNL) clusters and multi-GPU clusters as our target platforms. From an algorithm aspect, current distributed machine learning systems are mainly designed for cloud systems. These methods are asynchronous because of the slow network and high fault-tolerance requirement on cloud systems. We focus on Elastic Averaging SGD (EASGD) to design algorithms for HPC clusters. Original EASGD used round-robin method for communication and updating. The communication is ordered by the machine rank ID, which is inefficient on HPC clusters. First, we redesign four efficient algorithms for HPC systems to improve EASGD's poor scaling on clusters. Async EASGD, Async MEASGD, and Hogwild EASGD are faster \textcolor{black}{than} their existing counterparts (Async SGD, Async MSGD, and Hogwild SGD, resp.) in all the comparisons. Finally, we design Sync EASGD, which ties for the best performance among all the methods while being deterministic. In addition to the algorithmic improvements, we use some system-algorithm codesign techniques to scale up the algorithms. By reducing the percentage of communication from 87% to 14%, our Sync EASGD achieves 5.3x speedup over original EASGD on the same platform. We get 91.5% weak scaling efficiency on 4253 KNL cores, which is higher than the state-of-the-art implementation.

60 APPLIED LIFE SCIENCES↗

Versioned distributed arrays for resilience in scientific applications: Global view resilience

Exascale studies project reliability challenges for future high-performance computing (HPC) systems. We propose the Global View Resilience (GVR) system, a library that enables applications to add resilience in a portable, application-controlled fashion using versioned distributed arrays. We describe GVR’s interfaces to distributed arrays, versioning, and cross-layer error recovery. Using several large applications (OpenMC, the preconditioned conjugate gradient solver PCG, ddcMD, and Chombo), we evaluate the programmer effort to add resilience. The required changes are small (<2% LOC), localized, and machine-independent, requiring no software architecture changes. We also measure the overhead of adding GVR versioning and show that generally overheads <2% are achieved. We conclude that GVR’s interfaces and implementation are flexible and portable and create a gentle-slope path to tolerate growing error rates in future systems.

97 MATHEMATICS AND COMPUTING↗

Multi-Core Processor Memory Contention Benchmark Analysis Case Study

Multi-core processors dominate current mainframe, server, and high performance computing (HPC) systems. This paper provides synthetic kernel and natural benchmark results from an HPC system at the NASA Goddard Space Flight Center that illustrate the performance impacts of multi-core (dual- and quad-core) vs. single core processor systems. Analysis of processor design, application source code, and synthetic and natural test results all indicate that multi-core processors can suffer from significant memory subsystem contention compared to similar single-core processors.

Simon, Tyler↗

Preemptive resource management for dynamically arriving tasks in an oversubscribed heterogeneous computing system

We design resource management heuristics that assign serial tasks to the nodes of a heterogeneous high performance computing (HPC) system. The value of completing these tasks is modeled using monotonically decreasing utility functions that represent the time-varying importance of the task. The value of completing a task is equal to its utility function at the time of its completion. The overall performance of this system is measured using the total utility earned by all tasks during some interval of time. To maximize the performance of such a system where the preemption of tasks is possible, we have designed, analyzed, and compared a set of resource allocation heuristic techniques. We combine two utility-aware heuristics with three different preemption techniques to create six preemption-capable heuristics. We also consider the two utility-aware heuristics without preemption. We use simulation studies to evaluate this set of eight heuristics and compare them with an FCFS heuristic, which is often used in real systems, and random assignments. In general, our set of eight heuristics is able to significantly outperform the comparison heuristics, and the preemption-capable heuristics are able to significantly increase the utility earned compared to the heuristics that do not use preemption. We analyze the performance tradeoffs among the different preemption-capable heuristics under a variety of oversubscribed workload environments.

97 MATHEMATICS AND COMPUTING↗

A Roadmap for Edge Computing Enabled Automated Multidimensional Transmission Electron Microscopy

The advent of modern, high-speed electron detectors has made the collection of multidimensional hyperspectral transmission electron microscopy datasets, such as 4D-STEM, a routine. However, many microscopists find such experiments daunting since analysis, collection, long-term storage, and networking of such datasets remain challenging. Some common issues are their large and unwieldy size that often are several gigabytes, non-standardized data analysis routines, and a lack of clarity about the computing and network resources needed to utilize the electron microscope. The existing computing and networking bottlenecks introduce significant penalties in each step of these experiments, and thus, real-time analysis-driven automated experimentation for multidimensional TEM is challenging. One solution is to integrate microscopy with edge computing, where moderately powerful computational hardware performs the preliminary analysis before handing off the heavier computation to high-performance computing (HPC) systems. In this work, we trace the roots of computation in modern electron microscopy, demonstrate deep learning experiments running on an edge system, and discuss the networking requirements for tying together microscopes, edge computers, and HPC systems.

47 OTHER INSTRUMENTATION↗

A Quantum Analog Coprocessor for Correlated Electron Systems Simulation

Analog quantum simulation is an approach for studying physical systems that might otherwise be computationally intractable to simulate on classical high-performance computing (HPC) systems. The key idea behind analog quantum simulation is the realization of a physical system with a low-energy effective Hamiltonian that is the same as the low-energy effective Hamiltonian of some target system to be studied. Purpose-built nanoelectronic devices are a natural candidate for implementing the analog quantum simulation of strongly correlated materials that are otherwise challenging to study using classical HPC systems. However, realizing devices that are sufficiently large to study the properties of a non-trivial material system (e.g., those described by a Fermi-Hubbard model) will eventually require the fabrication, control, and measurement of at least 0(10) quantum dots, or other engineered quantum impurities. As a step toward large-scale analog or digital quantum simulation platforms based on nanoelectronic devices, we propose a new approach to analog quantum simulation that makes use of the large Hilbert space dimension of the electronic baths that are used to adjust the occupancy of one or a few engineered quantum impurities. This approach to analog quantum simulation allows us to study a wide array of quantum impurity models. We can further augment the computational power of such an approach by combining it with a classical computer to facilitate dynamical mean-field theory (DMFT) calculations. DMFT replaces the solution of a lattice impurity problem with the solution of a family of localized impurity problems with bath couplings that are adjusted to satisfy a self-consistency condition between the two models. In DMFT, the computationally challenging task is the high-accuracy solution of an instance of a quantum impurity model that is determined self-consistently in coordination with a mean-field calculation. We propose using one or a few engineered quantum impurities with adjustable couplings to baths to realize an analog quantum coprocessor that effects the solution of such a model through measurements of a physical quantum impurity, operating in coordination with a classical computer to achieve a self-consistent solution to a DMFT calculation. We focus on implementation details relevant to a number of technologies for which Sandia has design, fabrication, and measurement expertise. The primary technical advances outlined in this report concern the development of a supporting modeling capability. As with all analog quantum simulation platforms, the successful design and operation of individual devices depends critically on one's ability to predict the effective low-energy Hamiltonian governing its dynamics Our project has made this possible and lays the foundation for future experimental implementations.

71 CLASSICAL AND QUANTUM MECHANICS, GENERAL PHYSIC↗

CoREC: Scalable and Resilient In-memory Data Staging for In-situ Workflows

The dramatic increase in the scale of current and planned high-end HPC systems is leading new challenges, such as the growing costs of data movement and IO, and the reduced mean time between failures (MTBF) of system components. In-situ workflows, i.e., executing the entire application workflows on the HPC system, have emerged as an attractive approach to address data-related challenges by moving computations closer to the data, and staging-based frameworks have been effectively used to support in-situ workflows at scale. However, the resilience of these staging-based solutions has not been addressed, and they remain susceptible to expensive data failures. Furthermore, naive use of data resilience techniques such as n-way replication and erasure codes can impact latency and/or result in significant storage overheads. In this article, we present CoREC, a scalable and resilient in-memory data staging runtime for large-scale in-situ workflows. CoREC uses a novel hybrid approach that combines dynamic replication with erasure coding based on data access patterns. It also leverages multiple levels of replications and erasure coding to support diverse data resiliency requirements. Furthermore, the article presents optimizations for load balancing and conflict-avoiding encoding, and a low overhead, lazy data recovery scheme. We have implemented the CoREC runtime and have deployed with the DataSpaces staging service on leadership class computing machines and present an experimental evaluation in the article. Here, the experiments demonstrate that CoREC can tolerate in-memory data failures while maintaining low latency and sustaining high overall storage efficiency at large scales.

97 MATHEMATICS AND COMPUTING↗

Towards New Metrics for High-Performance Computing Resilience

Ensuring the reliability of applications is becoming an increasingly important challenge as high-performance computing (HPC) systems experience an ever-growing number of faults, errors and failures. While the HPC community has made substantial progress in developing various resilience solutions, it continues to rely on platform-based metrics to quantify application resiliency improvements. The resilience of an HPC application is concerned with the reliability of the application outcome as well as the fault handling efficiency. To understand the scope of impact, effective coverage and performance efficiency of existing and emerging resilience solutions, there is a need for new metrics. In this paper, we develop new ways to quantify resilience that consider both the reliability and the performance characteristics of the solutions from the perspective of HPC applications. As HPC systems continue to evolve in terms of scale and complexity, it is expected that applications will experience various types of faults, errors and failures, which will require applications to apply multiple resilience solutions across the system stack. The proposed metrics are intended to be useful for understanding the combined impact of these solutions on an application's ability to produce correct results and to evaluate their overall impact on an application's performance in the presence of various modes of faults.

Fault Tolerance↗

Award ER25750: Coordinated Infrastructure for Fault Tolerance Systems Indiana University Final Report

The main purpose of the Coordinated Infrastructure for Fault Tolerance in Systems initiative has been to conduct research with a goal of providing end-to-end fault tolerance on a systemwide basis for applications and other system software. While fault tolerance has been an integral part of most high-performance computing (HPC) system software developed over the past decade, it has been treated mostly as a collection of isolated stovepipes. Visibility and response to faults has typically been limited to the particular hardware and software subsystems in which they are initially observed. Little fault information is shared across subsystems, allowing little flexibility or control on a system-wide basis, making it practically impossible to provide cohesive end-to-end fault tolerance in support of scientific applications. As an example, consider faults such as communication link failures that can be seen by a network library but are not directly visible to the job scheduler, or consider faults related to node failures that can be detected by system monitoring software but are not inherently visible to the resource manager. If information about such faults could be shared by the network libraries or monitoring software, then other system software, such as a resource manager or job scheduler, could ensure that failed nodes or failed network links were excluded from further job allocations and that further diagnosis could be performed. As a founding member and one of the lead developers of the Open MPI project, our efforts over the course of this project have been focused on making Open MPI more robust to failures by supporting various fault tolerance techniques, and using fault information exchange and coordination between MPI and the HPC system software stack from the application, numeric libraries, and programming language runtime to other common system components such as jobs schedulers, resource managers, and monitoring tools.

97 MATHEMATICS AND COMPUTING↗

A Case For Intra-rack Resource Disaggregation in HPC

The expected halt of traditional technology scaling is motivating increased heterogeneity in high-performance computing (HPC) systems with the emergence of numerous specialized accelerators. As heterogeneity increases, so does the risk of underutilizing expensive hardware resources if we preserve today’s rigid node configuration and reservation strategies. This has sparked interest in resource disaggregation to enable finer-grain allocation of hardware resources to applications. However, there is currently no data-driven study of what range of disaggregation is appropriate in HPC. To that end, we perform a detailed analysis of key metrics sampled in NERSC’s Cori, a production HPC system that executes a diverse open-science HPC workload. In addition, we profile a variety of deep-learning applications to represent an emerging workload. We show that for a rack (cabinet) configuration and applications similar to Cori, a central processing unit with intra-rack disaggregation has a 99.5% probability to find all resources it requires inside its rack. In addition, ideal intra-rack resource disaggregation in Cori could reduce memory and NIC resources by 5.36% to 69.01% and still satisfy the worst-case average rack utilization.

97 MATHEMATICS AND COMPUTING↗

Implementing efficient data compression and encryption in a persistent key-value store for HPC

Recently, persistent data structures, like key-value stores (KVSs), which are stored in a high-performance computing (HPC) system’s nonvolatile memory, provide an attractive solution for a number of emerging challenges like limited I/O performance. Data compression and encryption are two well-known techniques for improving several properties of such data-oriented systems. This article investigates how to efficiently integrate data compression and encryption into persistent KVSs for HPC with the ultimate goal of hiding their costs and complexity in terms of performance and ease of use. Our compression technique exploits deep memory hierarchy in an HPC system to achieve both storage reduction and performance improvement. Our encryption technique provides a practical level of security and enables sharing of sensitive data securely in complex scientific workflows with nearly imperceptible cost. We implement the proposed techniques on top of a distributed embedded KVS to evaluate the benefits and costs of incorporating these capabilities along different points in the dataflow path, illustrating differences in effective bandwidth, latency, and additional computational expense on Swiss National Supercomputing Centre’s Grand Tavé and National Energy Research Scientific Computing Center’s Cori.

97 MATHEMATICS AND COMPUTING↗

Lightweight Provenance Service for High-Performance Computing

Provenance describes detailed information about the history of a piece of data, containing the relationships among elements such as users, processes, jobs, and workflows that contribute to the existence of data. Provenance is key to supporting many data management functionalities that are increasingly important in operations such as identifying data sources, parameters, or assumptions behind a given result; auditing data usage; or understanding details about how inputs are transformed into outputs. Despite its importance, however, provenance support is largely underdeveloped in highly parallel architectures and systems. One major challenge is the demanding requirements of providing provenance service in situ. The need to remain lightweight and to be always on often conflicts with the need to be transparent and offer an accurate catalog of details regarding the applications and systems. To tackle this challenge, we introduce a lightweight provenance service, called LPS, for high-performance computing (HPC) systems. LPS leverages a kernel instrument mechanism to achieve transparency and introduces representative execution and flexible granularity to capture comprehensive provenance with controllable overhead. Extensive evaluations and use cases have confirmed its efficiency and usability. We believe that LPS can be integrated into current and future HPC systems to support a variety of data management needs.

Dai, Dong↗

MaDaTS: Managing Data on Tiered Storage for Scientific Workflows

Scientific workflows are processing large amounts of data through complex simulation and analysis tasks. Meanwhile, the need to minimize I/O costs on next generation systems and the evolution of new technologies (NVRAMs, SSDs etc.) is resulting in deeper storage hierarchies on High Performance Computing (HPC) systems. A multi-tiered storage hierarchy introduces complexities in workflow and data management. There is need for simple and flexible data abstractions that can allow users to seamlessly manage workflow data and tasks on HPC systems with multiple storage tiers. MaDaTS (Managing Data on Tiered Storage for Scientific Workflows) provides an API and a command-line tool that allows users to manage their workflows and data on tiered storage (Ghoshal & Ramakrishnan (2017)).

97 MATHEMATICS AND COMPUTING↗

Accelerating Science Driven System Design With RAMP

Researchers from UC Berkeley, in collaboration with the Lawrence Berkeley National Lab, are engaged in developing an Infrastructure for Synthesis with Integrated Simulation (ISIS). The ISIS Project was a cooperative effort for “application-driven hardware design” that engages application scientists in the early parts of the hardware design process for future generation supercomputing systems. This project served to foster development of computing systems that are better tuned to the application requirements of demanding scientific applications and result in more cost-effective and efficient HPC system designs. In order to overcome long conventional design-cycle times, we leveraged reconfigurable devices to aid in the design of high-efficiency systems, including conventional multi- and many-core systems. The resulting system emulation/prototyping environment, in conjunction with the appropriate intermediate abstractions, provided both a convenient user programming experience and retained flexibility, and thus efficiency, of a reconfigurable platform. We initially targeted the Berkeley RAMP system (Research Accelerator for Multiple Processors) as that hardware emulation environment to facilitate and ultimately accelerate the iterative process of science-driven system design. Our goal was to develop and demonstrate a design methodology for domain-optimized computer system architectures. The tangible outcome is a methodology and tools for rapid prototyping and design-space exploration, leading to highly optimized and efficient HPC systems.

97 MATHEMATICS AND COMPUTING↗

Prediction and characterization of application power use in a high-performance computing environment

Power use in data centers and high-performance computing (HPC) facilities has grown in tandem with increases in the size and number of these facilities. Substantial innovation is needed to enable meaningful reduction in energy footprints in leadership-class HPC systems. In this paper, we focus on characterizing and investigating application-level power usage. We demonstrate potential methods for predicting power usage based on a priori and in situ characteristics. Lastly, we highlight a potential use case of this method through a simulated power-aware scheduler using historical jobs from a real scientific HPC system.

97 MATHEMATICS AND COMPUTING↗

Collective input/output under memory constraints

Compared with current high-performance computing (HPC) systems, exascale systems are expected to have much less memory per node, which can significantly reduce necessary collective input/output (I/O) performance. In this study, we introduce a memory-conscious collective I/O strategy that takes into account memory capacity and bandwidth constraints. The new strategy restricts aggregation data traffic within disjointed subgroups, coordinates I/O accesses in intranode and internode layers, and determines I/O aggregators at run time considering memory consumption among processes. We have prototyped the design and evaluated it with commonly used benchmarks to verify its potential. The evaluation results demonstrate that this strategy holds promise in mitigating the memory pressure, alleviating the contention for memory bandwidth, and improving the I/O performance for projected extreme-scale systems. Given the importance of supporting increasingly data-intensive workloads and projected memory constraints on increasingly larger scale HPC systems, this new memory-conscious collective I/O can have a significant positive impact on scientific discovery productivity.

collective input/output↗