Proceedings of the 47th International Conference on Parallel Processing Companion最新文献

GAS 气体

Proceedings of the 47th International Conference on Parallel Processing Companion Pub Date : 2018-08-13 DOI: 10.1145/3229710.3229758

D. Janakiram, S. Balaji, Akshay Dhumal, Nishank Garg, Ganesh Kulkarni

引用次数: 1

DVM 数字式电压表

Proceedings of the 47th International Conference on Parallel Processing Companion Pub Date : 2018-08-13 DOI: 10.1145/3229710.3229737

Abdullah Al-Mamun, Ke Wang, Jialin Liu, Dongfang Zhao

引用次数: 0

DROM

Proceedings of the 47th International Conference on Parallel Processing Companion Pub Date : 2018-08-13 DOI: 10.1145/3229710.3229752

Marco D'Amico, M. Garcia-Gasulla, Víctor López, Ana Jokanovic, R. Sirvent, J. Corbalán

{"title":"DROM","authors":"Marco D'Amico, M. Garcia-Gasulla, Víctor López, Ana Jokanovic, R. Sirvent, J. Corbalán","doi":"10.1145/3229710.3229752","DOIUrl":"https://doi.org/10.1145/3229710.3229752","url":null,"abstract":"In the design of future HPC systems, research in resource management is showing an increasing interest in a more dynamic control of the available resources. It has been proven that enabling the jobs to change the number of computing resources at run time, i.e. their malleability, can significantly improve HPC system performance. However, job schedulers and applications typically do not support malleability due to the common belief that it introduces additional programming complexity and performance impact. This paper presents DROM, an interface that provides efficient malleability with no effort for program developers. The running application is enabled to adapt the number of threads to the number of assigned computing resources in a completely transparent way to the user through the integration of DROM with standard programming models, such as OpenMP/OmpSs, and MPI. We designed the APIs to be easily used by any programming model, application and job scheduler or resource manager. Our experimental results from two realistic use cases analysis, based on malleability by reducing the number of cores a job is using per node and jobs co-allocation, show the potential of DROM for improving the performance of HPC systems. In particular, the workload of two MPI+OpenMP neuro-simulators are tested, reporting improvement in system metrics, such as total run time and average response time, up to 8% and 48%, respectively.","PeriodicalId":378200,"journal":{"name":"Proceedings of the 47th International Conference on Parallel Processing Companion","volume":"57 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114039166","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 12

iPregel

Proceedings of the 47th International Conference on Parallel Processing Companion Pub Date : 2018-08-13 DOI: 10.1145/3229710.3229719

L. Capelli, Zhenjiang Hu, Timothy A. K. Zakian

{"title":"iPregel","authors":"L. Capelli, Zhenjiang Hu, Timothy A. K. Zakian","doi":"10.1145/3229710.3229719","DOIUrl":"https://doi.org/10.1145/3229710.3229719","url":null,"abstract":"The expressiveness of the vertex-centric programming model introduced by Pregel attracted great attention. Over the years, numerous frameworks emerged, abiding by the same programming model, while relying on widely different architectural designs. The vast majority of existing vertex-centric frameworks exploits distributed memory parallelism or out-of-core computations. To our knowledge, only one vertex-centric framework is designed upon in-memory storage and shared memory parallelism. Unfortunately, while built on a faster architecture than that of other vertex-centric frameworks, it did not prove to significantly outperform other existing solutions. In this paper we present iPregel: another in-memory shared memory vertex-centric framework. The optimisations developed and presented in this paper particularly target three hotspots of vertex-centric calculations: selecting active vertices, routing messages to their recipient and updating recipients inbox. We compare iPregel against the state-of-the-art in-memory distributed memory framework Pregel+ on three of the most common vertex-centric applications: PageRank, Hashmin and the Single-Source Shortest Path. Experiments demonstrate that the single-node framework iPregel is faster than its distributed memory counterpart until at least 11 nodes are used. Further experiments show that iPregel completes a PageRank application with an order of magnitude less memory than popular vertex-centric frameworks.","PeriodicalId":378200,"journal":{"name":"Proceedings of the 47th International Conference on Parallel Processing Companion","volume":"240 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123258379","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 4

BAUSPACE

Proceedings of the 47th International Conference on Parallel Processing Companion Pub Date : 2018-08-13 DOI: 10.1145/3229710.3229744

D. Hernandez, Francisco Arcas-Túnez, Andrés Muñoz, José M. Cecilia

{"title":"BAUSPACE","authors":"D. Hernandez, Francisco Arcas-Túnez, Andrés Muñoz, José M. Cecilia","doi":"10.1145/3229710.3229744","DOIUrl":"https://doi.org/10.1145/3229710.3229744","url":null,"abstract":"The Internet of Things (IoT) is driving the next economic revolution where the main actors are both the data volume and the immediacy. However, the IoT world is increasingly generating vast amounts of data classified as a \"dark data\", since most of them are generated but never analysed. Therefore, efficient big data analysis in IoT infrastructure is becoming mandatory to transform this data deluge into meaningful information. Even after enabling this analysis, the quantitative information provided by traditional \"hard\" sensors is not enough to deal with some scenarios where human observations are required. These observations could be targeted through \"soft sensors\", where people's opinion in social networks, posts, news or comments may be analyzed to create dynamic observation resources. Combining both sources-of-information (devices and humans) automatically would provide a very powerful tool that could represent a step forward in the data understanding science. However, the development of soft sensors implies the use of many services for crawling text sources and mashup Web-based content, storage it, understanding the language or inferring information, just to mention a few. Therefore, a novel cloud-based distributed system is mandatory to be able to develop such frameworks. In this paper we introduce a work-in-progress for a distributed and modular framework to develop soft sensors in a scalable manner and transparently to the cloud provider.","PeriodicalId":378200,"journal":{"name":"Proceedings of the 47th International Conference on Parallel Processing Companion","volume":"16 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2018-08-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"123419371","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 2