OpenCL을 이용한 돈사 감시 응용의 효율적인 태스크 분배
Efficient Task Distribution for Pig Monitoring Applications Using OpenCL
0
조회
0
다운로드
3
인용
0
참고문헌
키워드
초록
다수의 태스크로 구성된 돈사 감시 응용은 내재된 데이터 병렬성을 활용하고 성능가속기를 사용하여 병렬 처리가 가능하다. 본 논문에서는 멀티코어 CPU와 매니코어 GPU로 구성된 이기종 컴퓨팅 플랫폼에서 돈사 감시 응용 수행 시 태스크 분배 방법을 제안한다. 즉, 각 태스크별로 OpenCL을 이용한 병렬 프로그램을 작성한 뒤, deviceCPU와 deviceGPU 각각에서 수행시켜 측정된 수행시간을 기준으로 가장 적합한 처리기를 결정한다. 제안 방법은 간단하지만 매우 효과적이고, CPU와 GPU로 구성된 이기종 컴퓨팅 플랫폼에서 다수의 태스크로 구성된 다른 응용을 병렬화하는 경우에도 적용될 수 있다. 실험 결과, 상이한 이기종 컴퓨팅 플랫폼에서 최적의 태스크 분배로 수행한 경우가 전체 태스크들을 deviceGPU에서 수행한 GPU-only 방법에 비교하여 각각 2.7배, 8.7배, 2.7배 성능 개선이 되었음을 확인하였다.
Abstract
Pig monitoring applications consisting of many tasks can take advantage of inherent data parallelism and enable parallel processing using performance accelerators. In this paper, we propose a task distribution method for pig monitoring applications into a heterogenous computing platform consisting of a multicore-CPU and a manycore-GPU. That is, a parallel program written in OpenCL is developed, and then the most suitable processor is determined based on the measured execution time of each task. The proposed method is simple but very effective, and can be applied to parallelize other applications consisting of many tasks on a heterogeneous computing platform consisting of a CPU and a GPU. Experimental results show that the performance of the proposed task distribution method on three different heterogeneous computing platforms can improve the performance of the typical GPU-only method where every tasks are executed on a deviceGPU by a factor of 1.5, 8.7 and 2.7, respectively.
참고문헌 (18)
- J. Sanders, "CUDA by Example" (2011), Addison Wesley
- J. Stone, "OpenCL : A Parallel Programming Standard for Heterogeneous Computing Systems", Computing in Science and Engineering, 12(3), pp.66~73 (2010)
- D. Kaeli, "Heterogeneous Computing with OpenCL 2.0" (2015), Morgan Kaufmann
- R. Couturier, "Designing Scientific Applications on GPUs" (2013), CRC Press
- W. Wang, "Adaptive Scheduling for Parallel Tasks with QoS Satisfaction for Hybrid Cloud Environments", J. of Supercomputing, 66(2), pp.783~811 (2013)
- Y. Hao, "Parallel Task Scheduling under Multi-Clouds", 11(1), pp.39~60 (2017), TIIS
- F. Ramezani, "Evolutionary Algorithm-based Multi-Objective Task Scheduling Optimization Model in Cloud Environments", WWW-Internet & Web Info. Sys, 18(6), pp.1737~1757 (2015)
- 김희곤, "이기종 컴퓨팅 환경에서 OpenCL을 사용한 포토모자이크 응용의 효율적인 작업부하 분배", 정보처리학회논문지. 컴퓨터 및 통신시스템, 4(8), pp.245~252 (2015), 한국정보처리학회[DOI]
- S. Lee, "CPU-GPU Hybrid Computing for Feature Extraction from Video Frame", IEICE Electronics Express, 11(22), pp.1~8 (2014)
- 사재원, "시공간 정보를 이용한 근접 돼지의 영상 분할", 정보처리학회논문지. 소프트웨어 및 데이터 공학, 4(10), pp.473~478 (2015), 한국정보처리학회[DOI]
- S. Zuo, "An Index Algorithm for Tracking Pigs in Pigsty", Proc. of International Conference on Information Technology and Management Science, pp.797~803 (2014)
- J. Lee, "Automatic Recognition of Aggressive Behavior in Pigs by using a Kinect Depth Sensor", Sensors, 16(5) (2016)
- B. Shao, "A Real-Time Computer Vision Assessment and Control of Thermal Comfort for Group-Housed Pigs", Computers and Electronics in Agriculture, 62(1), pp.15~21 (2008)
- J. Kim, "Pig’s Room Background Removal using Texture Information", Proc. of Smart Media Fall Conference (2016)
- N. Otsu, "A Threshold Selection Method from Gray-Level Histograms", IEEE Tr. Systems, Man, and Cybernetics, 9(1), pp.62~66 (1979)
- J. Han, "Enhanced computer vision with microsoft kinect sensor : A review", IEEE Transactions on Cybernetics, 43(5), pp.1318~1334 (2013)
- A. M. Reza, "Realization of the contrast limited adaptive histogram equalization(CLAHE)for real-time image enhancement", The Journal of VLSI Signal Processing, 38(1), pp.35~44 (2004)
- I. Jung, "Parallel Connected-Component Labeling Algorithm for GPGPU Applications", Proc. of ISCIT, pp.1149~1153 (2010)