检索结果-内蒙古大学图书馆

proceedings of the 15th acm International Conference on Future and Sustainable Energy systems

作者： Grant Wilkins Srinivasan Keshav Richard Mortier Department of Computer Science and Technology University of Cambridge United Kingdom

ISBN: (纸本)9798400704802

Both the training and use of Large Language Models (LLMs) require large amounts of energy. Their increasing popularity, therefore, raises critical concerns regarding the energy efficiency and sustainability of data centers that host them. This paper addresses the challenge of reducing energy consumption in data centers running LLMs. We propose a hybrid data center model that uses a cost-based scheduling framework to dynamically allocate LLM tasks across hardware accelerators that differ in their energy efficiencies and computational capabilities. Specifically, our workload-aware strategy determines whether tasks are processed on energy-efficient processors or high-performance GPUs based on the number of input and output tokens in a query. Our analysis of a representative LLM dataset, finds that this hybrid strategy can reduce CPU+GPU energy consumption by 7.5% compared to a workload-unaware baseline.

关键词： artificial intelligence heterogeneous computing large language models sustainable computing

来源：评论

学校读者我要写书评

暂无评论

CEEMS: A Resource Manager Agnostic Energy and Emissions Monitoring Stack 24

CEEMS: A Resource Manager Agnostic Energy and Emissions Moni...

引用

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

作者： Mahendra Paipuri Institut du Développement et des Ressources en Informatique Scientifique (IDRIS) National Centre for Scientific Research (CNRS) Orsay France

ISBN: (纸本)9798350355543

With the rapid acceleration of ML/AI research in the last couple of years, the energy consumption of the Information and Communication Technology (ICT) domain has rapidly increased. As a major part of this energy consumption is due to users' workloads, it is evident that users need to be aware of the energy footprint of their applications. Compute Energy & Emissions Monitoring Stack (CEEMS) has been designed to address this issue. CEEMS can report energy consumption and equivalent emissions of user workloads in real time for HPC and cloud platforms alike. Besides CPU energy usage, it supports reporting energy usage of workloads on NVIDIA and AMD GPU accelerators. CEEMS has been built around the prominent open-source tools in the observability eco-system like Prometheus and Grafana. CEEMS has been designed to be extensible and it allows the Data Center (DC) operators to easily define the energy estimation rules of user workloads based on the underlying hardware. This paper explains the architectural overview of CEEMS, data sources that are used to measure energy usage and estimate equivalent emissions and potential use cases of CEEMS from operator and user perspectives. Finally, the paper will conclude by describing how CEEMS deployment on the Jean-Zay supercomputing platform is capable of monitoring more than 1400 nodes that have a daily job churn rate of around 20k jobs.

关键词： Energy measurement

来源：评论

学校读者我要写书评

暂无评论

Dynamic Clustering-based Sharding in Distributed Deduplication systems 8

Dynamic Clustering-based Sharding in Distributed Deduplicati...

引用

8th IEEE/acm International Workshop on Data analysis and Reduction for Big Scientific Data, DRBSD-8 2022

作者： Zhou, Peng Zou, Xiangyu Xia, Wen Harbin Institute of Technology ShenZhen China

ISBN: (纸本)9781665463379

Witha massive upsurge in data, combining dedu-plication with distributed storage continuously suffer from a low deduplication ratio when providing the corresponding through-put. It is because distributed storage requires sharding data on different nodes, while global deduplication needs eliminating re-dundancies in a unified view. In this paper, we present clustering-based sharding method, D-Shard, in distributed deduplication storage systems that leads to a comparable deduplication effi-ciency on a single system while supporting a high throughput. To achieve that, we try to make that duplicates are located at the same node as much as possible with two key techniques. First, using Dynamic K-Means approach to cluster super-blocks, then extracting every cluster center feature as the anchor point for sharding;Second, Construct a secondary deduplication index based on the Compact Hamming Index. If a super-block mapping to the middle of two anchor points across nodes, then super-blocks adjacent to the feature are re-deduplicated in the cluster offline. Currently, preliminary results show that super-block clustering is convergent, and routing strategy based on anchor points can achieve a higher deduplication ratio compared to the state-of-the-art approach and the throughput of system has been greatly improved. © 2022 IEEE.

关键词： Cloud storage

来源：评论

学校读者我要写书评

暂无评论

Transaction Scheduling: From Conflicts to Runtime Conflicts

引用

proceedings of the acm on Management of Data 2023年第1期1卷 1-26页

作者： Yang Cao Wenfei Fan Weijie Ou Rui Xie Wenyue Zhao University of Edinburgh Edinburgh United Kingdom Shenzhen Institute of Computing Sciences University of Edinburgh BDBC & Beihang University Shenzhen China Shenzhen Institute of Computing Sciences Shenzhen China

This paper studies how to improve the performance of main memory multicore OLTP systems for executing transactions with conflicts. A promising approach is to partition transaction workloads into mutually conflict-free clusters, and distribute the clusters to different cores for concurrent execution. We show that if transactions in each cluster are properly scheduled, transactions that are traditionally considered conflicting can be executed without conflicts at runtime. In light of this, we propose to schedule transactions and reduce runtime conflicts, instead of partitioning based on the conventional notion of conflicts. We formulate the transaction scheduling problem to minimize runtime conflicts, and show that the problem is NP-complete. This said, we develop an efficient scheduling algorithm to improve parallelism. Moreover, for transactions that are not packed in batches, we show that runtime conflict analysis also helps reduce conflict penalties, by proposing a proactive deferring method. Using standard and enhanced benchmarks, we show that on average our scheduling and proactive deferring methods improve the throughput of existing partitioners and concurrency control protocols by 131% and 109%, respectively, up to 294% and 152%.

关键词： concurrency control transaction processing transaction scheduling

来源：评论

学校读者我要写书评

暂无评论

Dynamic Graph Neural Networks-Based Alert Link Prediction for Online Service systems 23

Dynamic Graph Neural Networks-Based Alert Link Prediction fo...

引用

proceedings of the 38th IEEE/acm International Conference on Automated Software Engineering

作者： Yiru Chen Chenxi Zhang Zhen Dong Dingyu Yang Xin Peng Jiayu Ou Hong Yang Zheshun Wu Xiaojun Qu Wei Li School of Computer Science Fudan University China and Shanghai Key Laboratory of Data Science Fudan University China and Shanghai Collaborative Innovation Center of Intelligent Visual Computing China Alibaba Group China

ISBN: (纸本)9798350329964

A fault in large online service systems often triggers numerous alerts due to the complex business and component dependencies among services, which is known as "alert storm". In a short time, an online service system may generate a huge amount of alert data. This poses a challenge for on-call engineers to identify alerts that are associated with a system failure for root cause analysis. In this paper, we propose DyAlert, a dynamic graph neural networks-based approach for linking alerts that might be triggered by a same fault to reduce the burden of on-call engineers in the fault analysis. Our insight is that alerts are often triggered by alert propagation when a system failure occurs, e.g., alert a would lead to the occurrence of alert b. Whether two alerts should be linked depends on if one alert is triggered by the propagation of the other. Leveraging this insight, we design a dynamic graph (namely Alert-Metric Dynamic Graph) that describes the propagation process of alerts. Based on the dynamic graph, we train a neural networks-based model to predict alert links. We evaluate DyAlert with real-world data collected from an online service system running 85 business units and about 30,000 different services in a large enterprise. The results show that DyAlert is effective in predicting alert links and it outperforms the state-of-the-art approaches with an average increase of 0.259 in F1-score.

关键词： linked alerts

来源：评论

学校读者我要写书评

暂无评论

Performance analysis of Federated Learning in Orbital Edge computing 23

Performance Analysis of Federated Learning in Orbital Edge C...

引用

proceedings of the IEEE/acm 16th International Conference on Utility and Cloud computing

作者： Mohammad Reza Jabbarpour Bahman Javadi Philip Leong Rodrigo N. Calheiros David Boland Chris Butler Western Sydney University Sydney NSW Australia Western Sydney University Sydney Australia University Of Sydney Sydney Australia University of Sydney Sydney Australia AUCloud Melbourne Australia

ISBN: (纸本)9798400702341

Federated Learning (FL) is a promising solution for collaborative machine learning while respecting data privacy and locality. FL has been used in Low Earth Orbit (LEO) satellite constellations for different space applications including earth observation, navigation, and positioning. Orbital Edge computing (OEC) refers to the deployment of edge computing resources and data processing capabilities in space-based systems, enabling real-time data analysis and decision-making for remote and space-based applications. While there is existing research exploring the integration of federated learning in OEC, the influence of diverse factors such as space conditions, communication constraints, and machine learning models remains uncertain. This paper addresses this gap and presents a comprehensive performance analysis of FL methods in the unique and challenging setting of OEC. We consider model accuracy, training time, and power consumption as the performance metrics under different working conditions including IID and non-IID data distributions to analyse the performance of centralised and decentralised FL approaches. The experimental results demonstrate that although the asynchronous centralised FL method has high fluctuations in the accuracy curve, it is suitable for space applications in which power consumption and training time are two main factors. In addition, the number of sampled satellites for decentralised FL methods is an important parameter in non-IID data distribution. Moreover, increasing altitude can reduce the training time and increase the power consumption. This study enables us to highlight a number of performance challenges in OEC for further investigation.

关键词： low-earth orbit federated learning orbital edge computing energy consumption performance analysis

来源：评论

学校读者我要写书评

暂无评论

Exploring the Data on Dual Pandemics: How Racism and COVID-19 Have Impacted the computing Community 21

Exploring the Data on Dual Pandemics: How Racism and COVID-1...

引用

52nd acm Technical Symposium on Computer Science Education, SIGCSE 2021

作者： Brown, Quincy Dubow, Wendy Payton, Jamie Quinn, Beth A. Bizot, Betsy *** Upper MarlboroMD United States University of Colorado at Boulder BoulderCO United States Temple University PhiladelphiaPA United States University of Colorado at Boulder BoulderCO United States Computing Research Association WashingtonDC United States

ISBN: (纸本)9781450380621

As the COVID-19 pandemic began to unfold, reports of the impact on lives, public and private industry, and education systems began to emerge. Almost simultaneously, the long-standing pandemic of racism in the U.S. was called to the forefront. The policies, structures, and members of the computing community are not immune from the impacts of these dual pandemics. While we do not yet know the long-term effects on the computing community, several efforts have emerged to advance our understanding through data collection and analysis. This panel present preliminary findings from surveys conducted by multiple organizations that explore the impact of these dual pandemics on conditions of work and education in the computing community, with a focus on the impacts on populations that have been historically underrepresented in computing. The panel invites discussion about promising (and not-so-promising) practices that have emerged in response to these pandemics and next steps to support the computing community. © 2021 Owner/Author.

关键词： COVID-19

来源：评论

学校读者我要写书评

暂无评论

proceedings of 2021 24th International Conference on Soft computing and measurements, SCM 2021

Proceedings of 2021 24th International Conference on Soft Co...

引用

24th International Conference on Soft computing and measurements, SCM 2021

ISBN: (纸本)9781665439749

The proceedings contain 71 papers. The topics discussed include: efficient user inspection algorithm based on dual bloom filters oriented for blockchain data management systems;secure outsourcing of fuzzy linear regression in cloud computing;design of the afterburning chamber mock-up casing;researching the relationship between information efficiency and complexity of digital information processing devices;soft measurement of process improvement potential;neuro-fuzzy model of gas balance control;cryptography professional rival (CPR): a game designing model to learn cryptography;assessing the degree of the social media user's openness using an expert model based on the Bayesian network;comprehensive analysis of cyber-physical systems data;measurement virtualization technologies for intelligent information and measurement systems;and fuzzy assessment of the competitiveness of cloud software products.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Study on Differential Expression Genes in HCC Based on GEO Database 2021

Study on Differential Expression Genes in HCC Based on GEO D...

引用

2021 International Conference on Bioinformatics and Intelligent computing, BIC 2021

作者： Zhao, Yani Xie, Yongfang Chongqing Key Laboratory of Big Data for Bio Intelligence Chongqing University of Posts and Telecommunications Chongqing China

ISBN: (纸本)9781450390002

Background: Primary hepatocellular carcinoma (HCC) is is a complex malignant tumor with high mortality. To explore the pathogenesis of HCC, Based on the GEO database, bioinformatics analysis was carried out based on the GEO database in this study. Methods: The gene expression datasets: GSE121248 were downloaded from the GEO database for bioinformatics analysis. Geo 2R was used to identify differentially expressed genes (DEGs) with the predefined criterion (P © 2021 acm.

关键词： Database systems

来源：评论

学校读者我要写书评

暂无评论

2022 IEEE 4th International Conference on Power, Intelligent computing and systems, ICPICS 2022

2022 IEEE 4th International Conference on Power, Intelligent...

引用

4th IEEE International Conference on Power, Intelligent computing and systems, ICPICS 2022

ISBN: (纸本)9781665467735

The proceedings contain 191 papers. The topics discussed include: virtual inertia control strategy for PV using dc capacitive and electrochemical energy storage;liquid drip detection in power plant based on machine vision;intelligent recognition algorithm of specific pattern content information based on mobile terminal equipment;research on power communication defect diagnosis technology based on unsupervised learning;key technologies of high speed modulator for satellite data transmission;research on privacy fraud detection of logistic regression based on homomorphic encryption;application research on x-ray imaging technology for online detection of gas insulated switchgear;power transformer state identification method based on operational deflection shapes and visual measurement technology;research on identification method of potential abnormal station for miniature circuit breaker;trajectory tracking error optimization based on iterative learning;research and design of the intelligent bulk transfer laboratory interconnection;energy management cloud platform in softwarized energy internet;simulation and application of a new power stealing method based on the comparison of two types of transformers;lane line detection based on machine vision;an analysis method of GIS equipment fault causes based on online monitoring and joint diagnosis;and virtual prototype design of flexible exoskeleton based on pattern recognition technology.

关键词：

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：