检索结果-内蒙古大学图书馆

Initial Experiences with DAOS Object Storage on Aurora

Initial Experiences with DAOS Object Storage on Aurora

2024 Workshops of the International Conference for High Performance computing, Networking, Storage and Analysis, SC Workshops 2024

作者： Latham, Rob Ross, Robert B. Carns, Philip Snyder, Shane Harms, Kevin Velusamy, Kaushik Coffman, Paul Mcpheeters, Gordon Argonne National Laboratory Mathematics and Computer Science Division LemontIL60439 United States Argonne National Laboratory Argonne Leadership Computing Facility LemontIL60439 United States

ISBN: (纸本)9798350355543

The storage subsystem of the Aurora platform at Argonne National Laboratory offers the potential for unprecedented application I/O performance. Its hardware stack combines NVMe drives, persistent memory devices, a large collection of storage servers, and a unified RDMA-capable network fabric. Its software stack is based on the DAOS object storage system, thereby putting into practice decades of I/O research into minimizing overhead, maximizing access concurrency, and streamlining application interfaces. Taken together, these elements combine to present up to 230 PiB of projected storage capacity and up to 31 TiB/s of projected I/O throughput to applications executing on *** storage platform is unique, however, and presents a new set of challenges to practitioners. To unlock the full potential of this system, we must revisit fundamental issues such as hardware topology, network concurrency, and I/O APIs to understand their impact on real-world I/O performance. In this paper we present our initial experiences with the DAOS storage system on Aurora and characterize its sensitivity to these parameters at small scale. We discuss our initial impressions and offer recommendations for how to extract maximum effective performance from this class of storage system. © 2024 IEEE.

关键词： daos object store storage

来源：评论

学校读者我要写书评

暂无评论

Scalable and Consistent Graph Neural Networks for Distributed Mesh-based Data-driven Modeling

Scalable and Consistent Graph Neural Networks for Distribute...

引用

2024 Workshops of the International Conference for High Performance computing, Networking, Storage and Analysis, SC Workshops 2024

作者： Barwey, Shivam Balin, Riccardo Lusch, Bethany Patel, Saumil Balakrishnan, Ramesh Pal, Pinaki Maulik, Romit Vishwanath, Venkatram Transportation and Power Systems Argonne National Laboratory United States Leadership Computing Facility Argonne National Laboratory United States Argonne National Laboratory Computational Science Division United States Pennsylvania State University Information Sciences and Technology United States Mathematics and Computer Science Argonne National Laboratory United States

ISBN: (纸本)9798350355543

This work develops a distributed graph neural network (GNN) methodology for mesh-based modeling applications using a consistent neural message passing layer. As the name implies, the focus is on enabling scalable operations that satisfy physical consistency via halo nodes at sub-graph boundaries. Here, consistency refers to the fact that a GNN trained and evaluated on one rank (one large graph) is arithmetically equivalent to evaluations on multiple ranks (a partitioned graph). This concept is demonstrated by interfacing GNNs with NekRS, a GPU-capable exascale CFD solver developed at Argonne National Laboratory. It is shown how the NekRS mesh partitioning can be linked to the distributed GNN training and inference routines, resulting in a scalable mesh-based data-driven modeling workflow. We study the impact of consistency on the scalability of mesh-based GNNs, demonstrating efficient scaling in consistent GNNs for up to O(1B) graph nodes on the Frontier exascale supercomputer. © 2024 IEEE.

关键词： graph neural networks mesh-based modeling

来源：评论

学校读者我要写书评

暂无评论

Initial Experiences with DAOS Object Storage on Aurora 24

Initial Experiences with DAOS Object Storage on Aurora

引用

Proceedings of the SC '24 Workshops of the International Conference on High Performance computing, Network, Storage, and Analysis

作者： Rob Latham Robert B. Ross Philip Carns Shane Snyder Kevin Harms Kaushik Velusamy Paul Coffman Gordon McPheeters Mathematics and Computer Science Division Argonne National Laboratory Lemont IL USA Argonne Leadership Computing Facility Argonne National Laboratory Lemont IL USA

ISBN: (纸本)9798350355543

关键词：

来源：评论

学校读者我要写书评

暂无评论

Initial Experiences with DAOS Object Storage on Aurora

Initial Experiences with DAOS Object Storage on Aurora

引用

High Performance computing, Networking, Storage and Analysis, SC-W: Workshops of the International Conference for

ISBN: (数字)9798350355543

ISBN: (纸本)9798350355550

关键词： Concurrent computing Performance evaluation Ion radiation effects Sensitivity Nonvolatile memory Throughput Hardware Topology Servers Magnetosphere

来源：评论

学校读者我要写书评

暂无评论

Protein Generation via Genome-scale Language Models with Bio-physical Scoring

Protein Generation via Genome-scale Language Models with Bio...

引用

2023 International Conference on High Performance computing, Network, Storage, and Analysis, SC Workshops 2023

作者： Dharuman, Gautham Ward, Logan Ma, Heng Setty, Priyanka V. Gokdemir, Ozan Foreman, Sam Emani, Murali Hippe, Kyle Brace, Alexander Keipert, Kristopher Gibbs, Thomas Foster, Ian Anandkumar, Anima Vishwanath, Venkatram Ramanathan, Arvind Data Science and Learning Division United States Argonne Leadership Computing Facility Argonne National Laboratory United States Computer Science Department University of Chicago United States Nvidia Inc. United States California Institute of Technology United States

ISBN: (纸本)9798400707858

Large language models (LLMs) trained on vast biological datasets can learn biological motifs and correlations across the evolutionary landscape of natural proteins. LLMs can then be used for de novo design of novel proteins with specific structures, functions, and physicochemical properties. We employ a pre-trained genome-scale language model that uses codons as tokens and integrate it into a workflow for targeted generation of sequences. Our framework suggests new gene sequences that are ranked for downstream evaluation by metrics that collectively capture extensive sequence-specific, biophysical, and biochemical properties. We demonstrate our integrated workflow to design novel variants of the enzyme, malate dehydrogenase (MDH), that exhibit more favorable activation energies than their natural counterparts (reduction of 4.01 kJ/mol) with sustained sequence generation rates of ~104/hr and simulation rates of ~102/hr on 64 nodes of Polaris with about 99.7 system utilization during the run. © 2023 ACM.

关键词： Physicochemical properties

来源：评论

学校读者我要写书评

暂无评论

Autotuning PolyBench benchmarks with LLVM Clang/Polly loop optimization pragmas using Bayesian optimization

Autotuning PolyBench benchmarks with LLVM Clang/Polly loop o...

引用

作者： Wu, Xingfu Kruse, Michael Balaprakash, Prasanna Finkel, Hal Hovland, Paul Taylor, Valerie Hall, Mary Mathematics & Computer Science Division Argonne National Laboratory LemontIL United States Argonne Leadership Computing Facility Argonne National Laboratory LemontIL United States Department of Computer Science University of Utah Salt Lake CityUT United States

We develop a ytopt autotuning framework that leverages Bayesian optimization to explore the parameter space search and compare four different supervised learning methods within Bayesian optimization and evaluate their effectiveness. We select six of the most complex PolyBench benchmarks and apply the newly developed LLVM Clang/Polly loop optimization pragmas to the benchmarks to optimize them. We then use the autotuning framework to optimize the pragma parameters to improve their performance. The experimental results show that our autotuning approach outperforms the other compiling methods to provide the smallest execution time for the benchmarks syr2k, 3mm, heat-3d, lu, and covariance with two large datasets in 200 code evaluations for effectively searching the parameter spaces with up to 170,368 different configurations. We find that the Floyd–Warshall benchmark did not benefit from autotuning. To cope with this issue, we provide some compiler option solutions to improve the performance. Then we present loop autotuning without a user's knowledge using a simple mctree autotuning framework to further improve the performance of the Floyd–Warshall benchmark. We also extend the ytopt autotuning framework to tune a deep learning application. © 2021 John Wiley & Sons Ltd.

关键词： Large dataset

来源：评论

学校读者我要写书评

暂无评论

ANALYZING THE IMPACT OF CLIMATE CHANGE ON CRITICAL INFRASTRUCTURE FROM THE SCIENTIFIC LITERATURE: A WEAKLY SUPERVISED NLP APPROACH

arXiv

引用

arXiv 2023年

作者： Mallick, Tanwi Bergerson, Joshua David Verner, Duane R. Hutchison, John K. Levy, Leslie-Anne Balaprakash, Prasanna Mathematics and Computer Science Division Argonne National Laboratory LemontIL United States Decision and Infrastructure Sciences Division Argonne National Laboratory LemontIL United States Mathematics and Computer Science Division Argonne Leadership Computing Facility Argonne National Laboratory LemontIL United States

Natural language processing (NLP) is a promising approach for analyzing large volumes of climate-change and infrastructure-related scientific literature. However, best-in-practice NLP techniques require large collections of relevant documents (corpus). Furthermore, NLP techniques using machine learning and deep learning techniques require labels grouping the articles based on user-defined criteria for a significant subset of a corpus in order to train the supervised model. Even labeling a few hundred documents with human subject-matter experts is a time-consuming process. To expedite this process, we developed a weak supervision-based NLP approach that leverages semantic similarity between categories and documents to (i) establish a topic-specific corpus by subsetting a large-scale open-access corpus and (ii) generate category labels for the topic-specific corpus. The developed approach performs programmatic labeling with weak supervision by defining labeling functions to capture semantic similarity, rather than syntactic or rule-based similarity, between documents and categories. To support this approach, we developed climate hazard definitions based on several authoritative sources on climate change and used definitions developed by the U.S. Department of Homeland Security for National Critical Functions (NCFs) that encompass the functions provided by the critical infrastructure. By comparing the semantic similarity between documents in the Semantic Scholar Open Research Corpus (S2ORC), we established a large corpus of 17,136 research articles focused on both climate change hazards and NCFs as a subset of S2ORC. We use multiple text-embedding models as weak learners within the programmatic labeling technique. To speed up the labeling process, we scaled the weak supervision labeling across multiple GPUs on a computing cluster at the Argonne leadership computing facility. In comparison with a months-long process of subject-matter expert labeling, we assign category

关键词： Embeddings

来源：评论

学校读者我要写书评

暂无评论

QUALITY MEASURES FOR DYNAMIC GRAPH GENERATIVE MODELS

arXiv

引用

arXiv 2025年

作者： Hosseini, Ryien Simini, Filippo Vishwanath, Venkatram Willett, Rebecca Hoffmann, Henry Department of Computer Science University of Chicago United States Leadership Computing Facility Argonne National Laboratory Department of Statistics University of Chicago United States NSF-Simons National Institute for Theory and Mathematics in Biology

Deep generative models have recently achieved significant success in modeling graph data, including dynamic graphs, where topology and features evolve over time. However, unlike in vision and natural language domains, evaluating generative models for dynamic graphs is challenging due to the difficulty of visualizing their output, making quantitative metrics essential. In this work, we develop a new quality metric for evaluating generative models of dynamic graphs. Current metrics for dynamic graphs typically involve discretizing the continuous-evolution of graphs into static snapshots and then applying conventional graph similarity measures. This approach has several limitations: (a) it models temporally related events as i.i.d. samples, failing to capture the non-uniform evolution of dynamic graphs;(b) it lacks a unified measure that is sensitive to both features and topology;(c) it fails to provide a scalar metric, requiring multiple metrics without clear superiority;and (d) it requires explicitly instantiating each static snapshot, leading to impractical runtime demands that hinder evaluation at scale. We propose a novel metric based on the Johnson-Lindenstrauss lemma, applying random projections directly to dynamic graph data. This results in an expressive, scalar, and application-agnostic measure of dynamic graph similarity that overcomes the limitations of traditional methods. We also provide a comprehensive empirical evaluation of metrics for continuous-time dynamic graphs, demonstrating the effectiveness of our approach compared to existing methods. Our implementation is available at https://***/ryienh/jl-metric. Copyright © 2025, The Authors. All rights reserved.

关键词： Generative adversarial networks

来源：评论

学校读者我要写书评

暂无评论

Formalizing the generalization-forgetting trade-off in continual learning 21

Formalizing the generalization-forgetting trade-off in conti...

引用

Proceedings of the 35th International Conference on Neural Information Processing Systems

作者： R. Krishnan Prasanna Balaprakash Mathematics and Computer Science Division Mathematics and Computer Science Division and Leadership Computing Facility Argonne National Laboratory

ISBN: (纸本)9781713845393

We formulate the continual learning problem via dynamic programming and model the trade-off between catastrophic forgetting and generalization as a two-player sequential game. In this approach, player 1 maximizes the cost due to lack of generalization whereas player 2 minimizes the cost due to increased catastrophic forgetting. We show theoretically and experimentally that a balance point between the two players exists for each task and that this point is stable (once the balance is achieved, the two players stay at the balance point). Next, we introduce balanced continual learning (BCL), which is designed to attain balance between generalization and forgetting, and we empirically demonstrate that BCL is comparable to or better than the state of the art.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Multiscale Graph Neural Network Autoencoders for Interpretable Scientific Machine Learning

arXiv

引用

arXiv 2023年

作者： Barwey, Shivam Shankar, Varun Viswanathan, Venkatasubramanian Maulik, Romit Argonne Leadership Computing Facility Argonne National Laboratory United States Department of Mechanical Engineering Carnegie Mellon University United States Mathematics & Computer Science Division Argonne National Laboratory United States

The goal of this work is to address two limitations in autoencoder-based models: latent space interpretability and compatibility with unstructured meshes. This is accomplished here with the development of a novel graph neural network (GNN) autoencoding architecture with demonstrations on complex fluid flow applications. To address the first goal of interpretability, the GNN autoencoder achieves reduction in the number nodes in the encoding stage through an adaptive graph reduction procedure. This reduction procedure essentially amounts to flowfield-conditioned node sampling and sensor identification, and produces interpretable latent graph representations tailored to the flowfield reconstruction task in the form of so-called masked fields. These masked fields allow the user to (a) visualize where in physical space a given latent graph is active, and (b) interpret the time-evolution of the latent graph connectivity in accordance with the time-evolution of unsteady flow features (e.g. recirculation zones, shear layers) in the domain. To address the goal of unstructured mesh compatibility, the autoencoding architecture utilizes a series of multi-scale message passing (MMP) layers, each of which models information exchange among node neighborhoods at various lengthscales. The MMP layer, which augments standard single-scale message passing with learnable coarsening operations, allows the decoder to more efficiently reconstruct the flowfield from the identified regions in the masked fields. Analysis of latent graphs produced by the autoencoder for various model settings are conducted using unstructured snapshot data sourced from large-eddy simulations in a backward-facing step (BFS) flow configuration with an OpenFOAM-based flow solver at high Reynolds numbers. Copyright © 2023, The Authors. All rights reserved.

关键词： Message passing

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：