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 37 records · Page 2

Toward Transparent Data Management in Multi-layer Storage Hierarchy for HPC Systems

Upcoming exascale high performance computing (HPC) systems are expected to comprise multi-tier storage hierarchy, and thus will necessitate innovative storage and I/O mechanisms. Traditional disk and block-based interfaces and file systems face severe challenges in utilizing capabilities of storage hierarchies due to the lack of hierarchy support and semantic interfaces. Object-based and semantically-rich data abstractions for scientific data management on large scale systems offer a sustainable solution to these challenges. Such data abstractions can also simplify users involvement in data movement. Here, we take the first steps of realizing such an object abstraction and explore storage mechanisms for these objects to enhance I/O performance, especially for scientific applications. We explore how an object-based interface can facilitate next generation scalable computing systems by presenting the mapping of data I/O from two real world HPC scientific use cases: a plasma physics simulation code (VPIC) and a cosmology simulation code (HACC). Our storage model stores data objects in different physical organizations to support data movement across layers of memory/storage hierarchy. Our implementation sclaes well to 16K parallel processes, and compared to the state of the art, such as MPI-IO and HDF5, our object-based data abstractions and data placement strategy in multi-level storage hierarchy achieves up to 7 X I/O performance improvement for scientific data.

97 MATHEMATICS AND COMPUTING↗

Ten million and one penguins, or, lessons learned from booting millions of virtual machines on HPC systems.

In this paper we describe Megatux, a set of tools we are developing for rapid provisioning of millions of virtual machines and controlling and monitoring them, as well as what we've learned from booting one million Linux virtual machines on the Thunderbird (4660 nodes) and 550,000 Linux virtual machines on the Hyperion (1024 nodes) clusters. As might be expected, our tools use hierarchical structures. In contrast to existing HPC systems, our tools do not require perfect hardware; that all systems be booted at the same time; and static configuration files that define the role of each node. While we believe these tools will be useful for future HPC systems, we are using them today to construct botnets. Botnets have been in the news recently, as discoveries of their scale (millions of infected machines for even a single botnet) and their reach (global) and their impact on organizations (devastating in financial costs and time lost to recovery) have become more apparent. A distinguishing feature of botnets is their emergent behavior: fairly simple operational rule sets can result in behavior that cannot be predicted. In general, there is no reducible understanding of how a large network will behave ahead of 'running it'. 'Running it' means observing the actual network in operation or simulating/emulating it. Unfortunately, this behavior is only seen at scale, i.e. when at minimum 10s of thousands of machines are infected. To add to the problem, botnets typically change at least 11% of the machines they are using in any given week, and this changing population is an integral part of their behavior. The use of virtual machines to assist in the forensics of malware is not new to the cyber security world. Reverse engineering techniques often use virtual machines in combination with code debuggers. Nevertheless, this task largely remains a manual process to get past code obfuscation and is inherently slow. As part of our cyber security work at Sandia National Laboratories, we are striving to understand the global network behavior of botnets. We are planning to take existing botnets, as found in the wild, and run them on HPC systems. We have turned to HPC systems to support the creation and operation of millions of Linux virtual machines as a means of observing the interaction of the botnet and other noninfected hosts. We started out using traditional HPC tools, but these tools are designed for a much smaller scale, typically topping out at one to ten thousand machines. HPC programming libraries and tools also assume complete connectivity between all nodes, with the attendant configuration files and data structures to match; this assumption holds up very poorly on systems with millions of nodes.

99 GENERAL AND MISCELLANEOUS↗

Analyzing File Access Patterns on Large-Scale HPC Systems: Opportunities for File Prefetching

This paper explores the potential opportunities for implementing file prefetching techniques on large-scale high-performance computing (HPC) systems. Specifically, we investigate the file access patterns of various applications across multiple scientific domains using two years' worth of Darshan I/O traces obtained from the Summit supercomputer. We identify recurring trends and patterns which indicate that prefetching can be effectively leveraged to improve data access performance on HPC systems. This study serves as a valuable reference for system architects and developers in the HPC community, providing insights into the opportunities and challenges associated with enabling file prefetching on large-scale HPC systems.

Karimi, Ahmad Maroof↗

Towards Interactive, Reproducible Analytics at Scale on HPC Systems

The growth in scientific data volumes has resulted in a need to scale up processing and analysis pipelines using High Performance Computing (HPC) systems. These workflows need interactive, reproducible analytics at scale. The Jupyter platform provides core capabilities for interactivity but was not designed for HPC systems. In this paper, we outline our efforts that bring together core technologies based on the Jupyter Platform to create interactive, reproducible analytics at scale on HPC systems. Our work is grounded in a real world science use case-applying geophysical simulations and inversions for imaging the subsurface. Our core platform addresses three key areas of the scientific analysis workflow-reproducibility, scalability, and interactivity. We describe our implemention of a system, using Binder, Science Capsule, and Dask software. We demonstrate the use of this software to run our use case and interactively visualize real-Time streams of HDF5 data.

containers↗

Design and implementation of I/O performance prediction scheme on HPC systems through large-scale log analysis

Abstract Large-scale high performance computing (HPC) systems typically consist of many thousands of CPUs and storage units used by hundreds to thousands of users simultaneously. Applications from large numbers of users have diverse characteristics, such as varying computation, communication, memory, and I/O intensity. A good understanding of the performance characteristics of each user application is important for job scheduling and resource provisioning. Among these performance characteristics, I/O performance is becoming increasingly important as data sizes rapidly increase and large-scale applications, such as simulation and model training, are widely adopted. However, predicting I/O performance is difficult because I/O systems are shared among all users and involve many layers of software and hardware stack, including the application, network interconnect, operating system, file system, and storage devices. Furthermore, updates to these layers and changes in system management policy can significantly alter the I/O behavior of applications and the entire system. To improve the prediction of the I/O performance on HPC systems, we propose integrating information from several different system logs and developing a regression-based approach to predict the I/O performance. Our proposed scheme can dynamically select the most relevant features from the log entries using various feature selection algorithms and scoring functions, and can automatically select the regression algorithm with the best accuracy for the prediction task. The evaluation results show that our proposed scheme can predict the write performance with up to 90% prediction accuracy and the read performance with up to 99% prediction accuracy using the real logs from the Cori supercomputer system at NERSC.

97 MATHEMATICS AND COMPUTING↗

Fault-Tolerant Deep Learning Cache with Hash Ring for Load Balancing in HPC Systems

Large-scale DL on HPC systems like Frontier and Summit uses distributed node-local caching to address scalability and performance challenges. However, as these systems grow more complex, the risk of node failures increases, and current caching approaches lack fault tolerance, jeopardizing large-scale training jobs. We analyzed six months of SLURM job logs from Frontier and found that over 30% of jobs failed after an average of 75 minutes. To address this, we propose fault-tolerance strategies that recache data lost from failed nodes using a hash ring technique for balanced data recaching in the distributed node-local caching, reducing reliance on the PFS. Our extensive evaluations on Frontier showed that the hash ring-based recaching approach reduced training time by approximately 25% compared to the approach that redirects I/O to the PFS after node failures and demonstrated effective load balancing of training data across nodes.

Lee, Seoyeong↗

Shrink or Substitute: Handling Process Failures in HPC Systems Using In-Situ Recovery

Efficient utilization of today's high-performance computing (HPC) systems with complex software and hardware components requires that the HPC applications are designed to tolerate process failures at runtime. With low mean-time-to-failure (MTTF) of current and future HPC systems, long running simulations on these systems requires capabilities for gracefully handling process failures by the applications themselves. In this paper, we explore the use of fault tolerance extensions to Message Passing Interface (MPI) called user-level failure mitigation (ULFM) for handling process failures without the need to discard the progress made by the application. We explore two alternative recovery strategies, which use ULFM along with application-driven in-memory checkpointing. In the first case, the application is recovered with only the surviving processes, and in the second case, spares are used to replace the failed processes, such that the original configuration of the application is restored. Our experimental results demonstrate that graceful degradation is a viable alternative for recovery in environments where spares may not be available.

Ashraf, Rizwan↗

I/O in Machine Learning Applications on HPC Systems: A 360-degree Survey

Growing interest in Artificial Intelligence (AI) has resulted in a surge in demand for faster methods of Machine Learning (ML) model training and inference. This demand for speed has prompted the use of high performance computing (HPC) systems that excel in managing distributed workloads. Because data is the main fuel for AI applications, the performance of the storage and I/O subsystem of HPC systems is critical. In the past, HPC applications accessed large portions of data written by simulations or experiments or ingested data for visualizations or analysis tasks. ML workloads perform small reads spread across a large number of random files. This shift of I/O access patterns poses several challenges to modern parallel storage systems. In this paper, we survey I/O in ML applications on HPC systems, and target literature within a 6-year time window from 2019 to 2024. We define the scope of the survey, provide an overview of the common phases of ML, review available profilers and benchmarks, examine the I/O patterns encountered during offline data preparation, training, and inference, and explore I/O optimizations utilized in modern ML frameworks and proposed in recent literature. Lastly, we seek to expose research gaps that could spawn further R&D.

97 MATHEMATICS AND COMPUTING↗

Trends in data locality abstractions for HPC systems

The cost of data movement has always been an important concern in high performance computing (HPC) systems. It has now become the dominant factor in terms of both energy consumption and performance. Support for expression of data locality has been explored in the past, but those efforts have had only modest success in being adopted in HPC applications for various reasons. However, with the increasing complexity of the memory hierarchy and higher parallelism in emerging HPC systems, locality management has acquired a new urgency. Developers can no longer limit themselves to low-level solutions and ignore the potential for productivity and performance portability obtained by using locality abstractions. Fortunately, the trend emerging in recent literature on the topic alleviates many of the concerns that got in the way of their adoption by application developers. Data locality abstractions are available in the forms of libraries, data structures, languages and runtime systems; a common theme is increasing productivity without sacrificing performance. Furthermore, this paper examines these trends and identifies commonalities that can combine various locality concepts to develop a comprehensive approach to expressing and managing data locality on future large-scale high-performance computing systems.

97 MATHEMATICS AND COMPUTING↗

Pin-pointing Node Failures in HPC Systems

Automated fault prediction and diagnosis in HPC systems needs to be efficient for better system resilience. With increasing scalability required for exascale, accurate fault prediction aiding in quick remedy is hard. With changing supercomputer architectures, distilling fault data from the noisy raw logs requires substantial efforts. Predicting node failures in such voluminous system logs is challenging. To this end, we investigate an interesting way to pin-point node failures in such supercomputing systems. Our study on Cray system data with automated machine learning tools suggests that specific patterns of event messages on node unavailability can be indicator to node failures. This data extraction coupled with system and job data correlation helps in devising a methodology to predict node failures and their location over a specific time frame. This work aims to enable broader applicability for a generic fault prediction framework.

Roman, E↗

Improving the Effectiveness of Burst Buffers for Big Data Processing in HPC Systems with Eley

Here, burst buffer is an effective solution for reducing the data transfer time and the I/O interference in HPC systems. Extending Burst Buffers (BBs) to handle Big Data applications is challenging because BBs must account for the large data inputs of Big Data applications and the Quality-of-Service (QoS) of HPC applications which are considered as first-class citizens in HPC systems. Existing BBs focus on only intermediate data of Big Data applications and incur a high performance degradation of both Big Data and HPC applications. We present Eley , a burst buffer solution that helps to accelerate the performance of Big Data applications while guaranteeing the QoS of HPC applications. To achieve this goal, Eley embraces interference-aware prefetching technique that makes reading data input faster while introducing low interference for HPC applications. Evaluations using a wide range of Big Data and HPC applications demonstrate that Eley improves the performance of Big Data applications by up to 30% compared to existing BBs while maintaining the QoS of HPC applications.

97 MATHEMATICS AND COMPUTING↗

DART: distributed adaptive radix tree for efficient affix-based keyword search on HPC systems

© 2018 Association for Computing Machinery. Affix-based search is a fundamental functionality for storage systems. It allows users to find desired datasets, where attributes of a dataset match an affix. While building inverted index to facilitate efficient affix based keyword search is a common practice for standalone databases and for desktop file systems, building local indexes or adopting indexing techniques used in a standalone data store is insufficient for highperformance computing (HPC) systems due to the massive amount of data and distributed nature of the storage devices within a system. In this paper, we propose Distributed Adaptive Radix Tree (DART), to address the challenge of distributed affix-based keyword search on HPC systems. This trie-based approach is scalable in achieving efficient affix-based search and alleviating imbalanced keyword distribution and excessive requests on keywords at scale. Our evaluation at different scales shows that, comparing with the "full string hashing" use case of the most popular distributed indexing technique - Distributed Hash Table (DHT), DART achieves up to 55× better throughput with prefix search and with suffix search, while achieving comparable throughput with exact and infix searches. Also, comparing to the "initial hashing" use case of DHT, DART maintains a balanced keyword distribution on distributed nodes and alleviates excessive query workload against popular keywords.

Zhang, Wei↗

Predicting runtime and resource utilization of jobs on integrated cloud and HPC systems

Recent advances in virtualization technologies used in cloud computing offer performance that closely approaches bare-metal levels. Combined with specialized instance types and high-speed networking services for cluster computing, cloud platforms have become a compelling option for high-performance computing (HPC). However, most current batch job schedulers in HPC systems are designed for homogeneous clusters and make decisions based on limited information about jobs and system status. Scientists typically submit computational jobs to these schedulers with a requested runtime that is often over- or under-estimated. More accurate runtime predictions can help schedulers make better decisions and reduce job turnaround times. Here, they can also support decisions about migrating jobs to the cloud to avoid long queue wait times in HPC systems.

97 MATHEMATICS AND COMPUTING↗

Towards Native Execution of Deep Learning on a Leadership-Class HPC System

Large parallel machines generally offer the best parallel performance with "native execution" that is achieved using codes developed with the optimized compilers, communication libraries, and runtimes offered on the machines. In this paper, we report and analyze performance results from native execution of deep learning on a leadership-class high-performance computing (HPC) system. Using our new code called DeepEx, we present a study of the parallel speed up and convergence rates of learning achieved with native parallel execution. In the trade-off between computational parallelism and synchronized convergence, we first focus on maximizing parallelism while still obtaining convergence. Scaling results are reported from execution on up to 15,000 GPUs using two scientific data sets from atom microscopy and protein folding applications, and also using the popular ImageNet data set. In terms of the traditional measure of parallel speed up, excellent scaling is observed up to 12,000 GPUs. Additionally, accounting for convergence rates of deep learning accuracy or error, a deep learning-specific metric called "learning speed up" is also tracked. The performance results indicate the need to evaluate parallel deep learning execution in terms of learning speed up, and point to additional directions for improved exploitation of high-end HPC systems.

Yoginath, Srikanth↗

SPEChpc 2021 Benchmark Suites for Modern HPC Systems

The SPEChpc 2021 suites are application-based benchmarks de- signed to measure performance of modern HPC systems. The bench- marks support MPI, MPI+OpenMP, MPI+OpenMP target offload, MPI+OpenACC and are portable across all major HPC platforms.

Boehm, Swen↗

Energy–Performance Trade-offs in Privacy-Preserving Federated Learning on SmartNIC-Enabled HPC Systems

Federated learning (FL) is increasingly deployed on accelerator-rich high-performance computing (HPC) systems, yet the system-level energy cost of privacy-aware FL remains poorly understood, particularly across heterogeneous networking and server-placement options. We present a measurement-driven study of energy–performance trade-offs for FL on GH200-class nodes across three deployment configurations: CPU-Ethernet, CPU-InfiniBand (RDMA-capable), and a DPU-hosted FL server over InfiniBand using a BlueField-3 SmartNIC/DPU. Using NVIDIA FLARE (NVFLARE), we align node-level power telemetry with per-round timing extracted from NVFLARE logs to quantify time-to-solution (TTS), energy-to-solution (ETS), energy-delay product (EDP), and synchronization behavior for three transformer models (ALBERT, DistilBERT, BERT), trained with and without differential privacy (DP). We find that interconnect choice is the dominant driver of runtime and energy: host-managed InfiniBand consistently reduces communication overhead versus Ethernet, yielding lower TTS/ETS/EDP. In contrast, in our NVFLARE deployment, placing the FL server on the DPU does not consistently match CPU-InfiniBand performance and can be slower—especially for larger models—highlighting that server placement alone is not sufficient to guarantee end-to-end gains. Finally, under our fixed-round protocol, DP increases per-round cost and runtime variance; ETS increases largely in proportion to TTS because average node power remains relatively stable across configurations.

Kotevska, Olivera [ORNL] (ORCID:0000000316772243)↗

Regen: An object layout regenerator on large-scale production HPC systems

This article proposes an object layout regenerator called Regen which regenerates and removes the object layout dynamically to improve the read performance of applications. Regen first detects frequent access patterns from the I/O requests of the applications. Second, Regen reorganizes the objects and regenerates or preallocates new object layouts according to the identified access patterns. Finally, Regen removes or reuses the obsolete or regenerated object layouts as necessary. As a result, Regen accelerates access to objects by providing a flexible object layout. We implement Regen as a framework on top of Proactive Data Container (PDC) and evaluate it on Cori supercomputer, a production-scale HPC system, by using realistic HPC I/O benchmarks. The experimental results show that Regen improves the I/O performance by up to 16.92 × compared with an existing system.

Distributed file system↗