Digital Library[ Search Result ]
Performance Analysis of Concurrent Multitasking for Efficient Resource Utilization of GPUs
Sejin Kim, Qichen Chen, HeonYoung Yeom, Yoonhee Kim
http://doi.org/10.5626/JOK.2021.48.6.604
As Graphics Processing Units (GPUs) are widely utilized to accelerate compute-intensive applications, their application has expanded especially in data centers and clouds. However, the existing resource sharing methods within GPU are limited and cannot efficiently handle several requests of concurrent cloud users’ executions on GPU while effectively utilizing the available system resources. In addition, it is challenging to effectively partition resources within GPU without understanding and assimilating application execution patterns. This paper proposes an execution pattern-based application classification method and analyzes run-time characteristics: why the performance of an application is saturated at a point regardless of the allocated resources. In addition, we analyze the multitasking performance of the co-allocated applications using smCompactor, a thread block-based scheduling framework. We identify near-best co-allocated application sets, which effectively utilize the available system resources. Based on our results, there was a performance improvement of approximately 28% compared to NVIDIA MPS.
An Execution Planning Scheme for Computational Applications using Profiling of a CPU-GPU Container Cluster
Jisun Oh, Sejin Kim, Yoonhee Kim
http://doi.org/10.5626/JOK.2019.46.10.975
As heterogeneous clusters and cloud environments have gained popularity, selection of appropriate resources in an integrated environment is considered essential for enhancement in application performance with reasonable resource utilization. Application characteristics may require specific resource in a certain order during its execution and ask smart job deployment among diverse nodes. Especially, it is necessary to have an execution plan in advance for better performance of CPU and GPU container clusters. In this paper, we propose an execution planning scheme based on runtime profiling history to place jobs on CPU-GPU nodes. Computational applications usually show good performance in GPU. However, the lack of GPU sharing methods leads to failure of co-locating jobs on a GPU node. Based on the profile information, the scheme provides the combination of applications to run at the same time in GPU container clusters and estimate the performance of workload before executing application among CPU-GPU container clusters. We have also demonstrated adjustment of the order of application execution using profile history in order to reduce the execution time of total workload by monitoring GPU memory usage.
Design of Video Advertisement Analysis via Analysis of Internet Term Sensitivity
Sejin Kim, Jieun Kim, Wonyoung Seong, Yoonhee Kim
http://doi.org/10.5626/JOK.2019.46.9.919
Analysis of the increasing influence of video advertisements via Social Networking Service (SNS) is important in identifying their effects. However, the traditional methods of survey-based analysis are not suitable for measurement of the effectiveness of SNS advertisements that are distributed rapidly via smartphone use and the current system does not consider the sensitivity of users expressed in various forms, such as slang, and emoticons. This study proposes an automated system for the analysis of the effects of video ads via video comments, reflecting the characteristics of short Korean sentences.
This system uses machine learning for the interpretation of Internet terms and compilation of a sentiment dictionary specializing in SNS short sentences. Emoticon, which is used to emphasize the sensitivity of users in comments, is used for sentiment analysis when applied to Korean syntax rules, and the system is designed and implemented for more sophisticated emotional analysis by calculating the emotional values of nouns that are subject to sentiment.
A Data Replacement Scheme considering Data Transfer in a Scientific Workflow and Change in the Experimental Environment
Julim Ahn, Heewon Kim, Yoonhee Kim
http://doi.org/10.5626/JOK.2018.45.12.1227
Scientific workflow applications have a large amount of data scattered in the data center, and as they use and execute applications, the execution results may vary depending on the location of the stored data. The location of the intermediate data produced during execution also affects the transmission. So, it is important for the location of the data to minimize the data transfer time and size. Therefore, we propose data placement considering the state of dynamically changing resources for data-intensive workflow applications. Considering the dynamically changing state of the resource during the execution of the task, the replacements in the data-intensive steps lead to a reduction in the data transfer time and the size of the transfer data.
Dynamic Memory Allocation for Scientific Workflows in Containers
Theodora Adufu, Jieun Choi, Yoonhee Kim
The workloads of large high-performance computing (HPC) scientific applications are steadily becoming “bursty” due to variable resource demands throughout their execution life-cycles. However, the over-provisioning of virtual resources for optimal performance during execution remains a key challenge in the scheduling of scientific HPC applications. While over-provisioning of virtual resources guarantees peak performance of scientific application in virtualized environments, it results in increased amounts of idle resources that are unavailable for use by other applications. Herein, we proposed a memory resource reconfiguration approach that allows the quick release of idle memory resources for new applications in OS-level virtualized systems, based on the applications resource-usage pattern profile data. We deployed a scientific workflow application in Docker, a light-weight OS-level virtualized system. In the proposed approach, memory allocation is fine-tuned to containers at each stage of the workflows execution life-cycle. Thus, overall memory resource utilization is improved.
A Design of a Distributed Computing Problem Solving Environment for Dietary Data Analysis
Jieun Choi, Younsun Ahn, Yoonhee Kim
Recently, wellness has become an issue related to improvements in personal health and quality of life. Data that are accumulated daily, such as meals and momentum records, in addition to body measurement information such as body weight, BMI and blood pressure have been used to analyze the personal health data of an individual. Therefore, it has become possible to prevent potential disease and to analyze dietary or exercise patterns. In terms of food and nutrition, analyses are performed to evaluate the health status of an individual using dietary data. However, it is very difficult to process the large amount of dietary data. An analysis of dietary data includes four steps, and each step contains a series of iterative tasks that are executed over a long time. This paper proposes a problem solving environment that automates dietary data analysis, and the proposed framework increases the speed with which an experiment can be conducted.
Search
Journal of KIISE
- ISSN : 2383-630X(Print)
- ISSN : 2383-6296(Electronic)
- KCI Accredited Journal
Editorial Office
- Tel. +82-2-588-9240
- Fax. +82-2-521-1352
- E-mail. chwoo@kiise.or.kr