Data-flow analysis and optimization for data coherence in heterogeneous architectures

Sousa, Rafael T. de; Pereira, Marcio Machado; Pereira, Fernando Magno Quintão; Araújo, Guido

doi:10.1016/j.jpdc.2019.04.004

Cited by 4 publications

(3 citation statements)

References 24 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Users put forward service requests to the cloud platform according to their own business needs. How to provide the fastest service with the minimum response time, ensure the effective access of users, and realize or eliminate invalid access is extremely important [7][8][9].…”

Section: Introductionmentioning

confidence: 99%

Virtual Simulation Management of Data Traffic Optimization of Big Data Cloud Platform considering Multipoint Mapping Algorithm

2021

Advances in Multimedia

View full text Add to dashboard Cite

With the continuous development of Internet, cloud computing, and other technologies, build a cloud platform based on Cloud Computing Center, but how to effectively carry out operation and maintenance and face users to ensure the continuity and effectiveness of the platform is extremely important. In view of these needs and limitations, this paper introduces the multipoint mapping algorithm, combs the statistical methods of platform cloud traffic, carries out platform data traffic by classification, constructs the data traffic optimization management model, analyzes the relevant data samples, carries out statistical calculation for data diversion tasks, analyzes and processes the priority indicators, and forms the final results through continuous iteration, realizing the management of data flow optimization virtual simulation of big data cloud platform. Simulation results show that the multipoint mapping algorithm is effective and can effectively support the data flow of big data cloud platform and optimize virtual simulation management.

show abstract

Section: Introductionmentioning

confidence: 99%

Virtual Simulation Management of Data Traffic Optimization of Big Data Cloud Platform considering Multipoint Mapping Algorithm

2021

Advances in Multimedia

View full text Add to dashboard Cite

show abstract

“…AMM [24] eliminates redundant transmissions, but is bound with X10CUDA [27]. Sousa et al [29] perform data coherence analysis on OpenCL code and then insert appropriate OpenCL function calls to minimize the number of data coherence operations. Our work also eliminates redundant transmissions, but cannot support multilevel pointers for the safety concern of mmap() function.…”

Section: Automatic Communication Management For Heterogeneous Computingmentioning

confidence: 99%

Automatic translation of data parallel programs for heterogeneous parallelism through OpenMP offloading

et al. 2020

View full text Add to dashboard Cite

show abstract

“…The paper Data-flow Analysis and Optimization for Data Coherence in Heterogeneous Architectures by Sousa, Pereira, Pereira and Araujo [8] proposes an innovative technique to allocate shared buffers between host and devices in OpenCL, that aims to minimize the number of data coherence operations. The experimental results showed speed-ups of up to 5.25x on an ARM Mali-T880 and up to 8.87x on an NVIDIA GPU Pascal Titan X.…”

Section: Special Issue Contentsmentioning

confidence: 99%

Trends on heterogeneous and innovative hardware and software systems

Melo

Carretero

Stenström

et al. 2019

Journal of Parallel and Distributed Computing

View full text Add to dashboard Cite

Data-flow analysis and optimization for data coherence in heterogeneous architectures

Cited by 4 publications

References 24 publications

Virtual Simulation Management of Data Traffic Optimization of Big Data Cloud Platform considering Multipoint Mapping Algorithm

Virtual Simulation Management of Data Traffic Optimization of Big Data Cloud Platform considering Multipoint Mapping Algorithm

Automatic translation of data parallel programs for heterogeneous parallelism through OpenMP offloading

Trends on heterogeneous and innovative hardware and software systems

Contact Info

Product

Resources

About