Performance Evaluation of Parallel Haemodynamic Computations on Heterogeneous Clouds

Bystrov, Oleg; Kačeniauskas, Arnas; Pacevič, Ruslan; Starikovičius, Vadimas; Maknickas, Algirdas; Stupak, E.; Igumenov, Aleksandr

doi:10.31577/cai_2020_4_695

Cited by 7 publications

(6 citation statements)

References 0 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…On average, Nova with KVM-based VMs outperformed Zun with Docker containers by 2.5% of the execution time on the native hardware. Growth in the infrastructure overhead, increasing the number of processes used to solve the aortic valve problem by a parallel SaaS based on the finite volume method and commercial ANSYS Fluent software, was also observed for Docker containers of the Open-Stack cloud [36]. The observed overhead values were less than 5% of the total execution time.…”

Section: Overhead Of Cloud Infrastructurementioning

confidence: 83%

“…Reddy and Lastovetsky [35] formulated a bi-objective optimization problem for performance and energy for data-parallel applications on homogeneous clusters. Bystrov et al [36] investigated a tradeoff between the computing speed and the consumed energy of a real-life hemodynamic application on a heterogeneous cloud. Parallel speedups obtained by using several domain decomposition methods were compared, but load balance and communication issues were not explored.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Performance of Communication- and Computation-Intensive SaaS on the OpenStack Cloud

2021

Self Cite

View full text Add to dashboard Cite

The pervasive use of cloud computing has led to many concerns, such as performance challenges in communication- and computation-intensive services on virtual cloud resources. Most evaluations of the infrastructural overhead are based on standard benchmarks. Therefore, the impact of communication issues and infrastructure services on the performance of parallel MPI-based computations remains unclear. This paper presents the performance analysis of communication- and computation-intensive software based on the discrete element method, which is deployed as a service (SaaS) on the OpenStack cloud. The performance measured on KVM-based virtual machines and Docker containers of the OpenStack cloud is compared with that obtained by using native hardware. The improved mapping of computations to multicore resources reduced the internode MPI communication by 34.4% and increased the parallel efficiency from 0.67 to 0.78, which shows the importance of communication issues. Increasing the number of parallel processes, the overhead of the cloud infrastructure increased to 13.7% and 11.2% of the software execution time on native hardware in the case of the Docker containers and KVM-based virtual machines of the OpenStack cloud, respectively. The observed overhead was mainly caused by OpenStack service processes that increased the load imbalance of parallel MPI-based SaaS.

show abstract

Section: Overhead Of Cloud Infrastructurementioning

confidence: 83%

Section: Introductionmentioning

confidence: 99%

Performance of Communication- and Computation-Intensive SaaS on the OpenStack Cloud

2021

Self Cite

View full text Add to dashboard Cite

show abstract

“…The diversity of cloud resources, followed by the trade‐off between cost and performance, makes resource selection a challenging task for cloud users in the case of parallel communication‐intensive applications. Bystrov et al 8 focus on the cost‐ and performance‐aware resource selection for the discrete element method (DEM) application based on the hybrid MPI + OpenCL parallelization scheme and deployed on a heterogeneous OpenStack cloud. The proposed resource selection method uses a preliminary application‐specific benchmark and the performance prediction based on the speedup of parallel computations to obtain Pareto optimal solutions and select the best configuration of containers.…”

Section: Accepted Papers For the Special Issue: Summarymentioning

confidence: 99%

Advances into exascale computing

Wyrzykowski,

Szymanski

2024

Concurrency and Computation

View full text Add to dashboard Cite

The landscape of high-performance computing (HPC) has been expanding with new technologies and increased system complexity. For hardware, this trend is driven by technological inventions increasing computing power capabilities while taming cost metrics. For applications, we are witnessing increasing growth in algorithms' complexity to accommodate constantly expanding data sizes and take full advantage of increasingly complex processing and storage characteristics. Successfully handling this expansion and maintaining performance and efficiency on adequate levels requires software that matches the emerging hardware and system innovations and can address concerns arising from evolving and new paradigms. An example of the evolving paradigms is heterogeneity-one of the most challenging properties of emerging parallel, distributed, and edge computing platforms. An appealing example of the new paradigms is a vigorous expansion of artificial intelligence (AI) and machine learning (ML) methods that have become pervasive in solving the most demanding problems across many science and engineering disciplines. Also, the approaching end of Moore's Law scaling requires exciting but complex innovations and advances in HPC hardware/software environments that need to be matched by advances in algorithms and software systems to meet the demands of the applications.The International Conference for Parallel Processing and Applied Mathematics (PPAM) has become, over the last 28 years, the premier forum for exchanging ideas between researchers involved in parallel and distributed computing, including theory and applications, as well as applied and computational mathematics. Thirteen previous events have been held in different universities in Poland since 1994, when the first PPAM took place in Czestochowa.The works presented at the PPAM 2022 covered topics demonstrating the increased diversity of the HPC ecosystem and the clear need for innovative solutions in this area. The wide range of topics includes enabling technologies for next-generation HPC systems, along with heterogeneous computing, quantum computing and communication; exploitation of parallel, distributed, and edge computing; algorithms for efficient problem solving on emerging HPC systems; programming paradigms and tools for HPC systems; scheduling and load balancing for parallel computing; modeling, characterization, and optimization for next-generation HPC systems; applications and software engineering for HPC platforms; methods and tools for the parallel solution of large-scale problems, including AI-and ML-based techniques. This special issue of Concurrency and Computation: Practice and Experience contains eight selected papers from the PPAM 2022 Conference.We hope you find them interesting and that they stimulate new ideas and future advancements in innovative technologies and models, algorithms, and software tools that facilitate efficient and convenient utilization of emerging HPC systems.

show abstract

“…Their proposed method tried its best finishing more jobs and minimizing the average waiting time. Bystrov et al 21 investigated a trade‐off between the computing speed and the consumed energy of a hemodynamic application on a heterogeneous OpenStack cloud. Parallel speedups obtained by using several domain decomposition methods were compared, but prediction of application performance was not employed for resource selection.…”

Section: The Related Workmentioning

confidence: 99%

Cost‐ and performance‐aware resource selection for parallel software on heterogeneous cloud

Bystrov

Pacevič

Kačeniauskas

2023

Concurrency and Computation

Self Cite

View full text Add to dashboard Cite

SummaryCloud providers offer flexible infrastructures and on‐demand services, including the capability to deploy low cost virtual resources of many different types. However, the diversity of cloud resources followed by the important trade‐off between cost and performance makes the resource selection a challenging task for users in the case of parallel communication‐intensive software. The paper presents cost‐ and performance‐aware resource selection for parallel discrete element method (DEM) software as a service (SaaS) on heterogeneous OpenStack cloud. The developed resource selection uses preliminary application‐specific benchmarks of size smaller than targeted problems and the performance prediction based on speedup of parallel computations to obtain Pareto optimal solutions and to select the best configuration of containers from user's perspective. Hybrid parallelization of DEM software is developed by using OpenCL for shared‐memory multi‐core architectures and MPI for internode communications on distributed‐memory computer clusters. Round up and proportional pricing schemes are examined and compared from a user's perspective. Lower cost of computations obtained by using the proportional pricing scheme is always preferable for users. However, the difference approaches 1.0% of the cost calculated by using proportional pricing scheme, when long lasting computations are performed. The prediction tends to underestimate the execution time of DEM SaaS, but its accuracy is sufficient to obtain the same Pareto optimal solutions by using measured and predicted execution times. Pareto front and linear scalarization propose to select configurations of containers capable of exploiting higher memory bandwidth, which is specific to memory bandwidth bound DEM computations.

show abstract

Performance Evaluation of Parallel Haemodynamic Computations on Heterogeneous Clouds

Cited by 7 publications

References 0 publications

Performance of Communication- and Computation-Intensive SaaS on the OpenStack Cloud

Performance of Communication- and Computation-Intensive SaaS on the OpenStack Cloud

Advances into exascale computing

Cost‐ and performance‐aware resource selection for parallel software on heterogeneous cloud

Contact Info

Product

Resources

About