엣지 컴퓨팅 환경에서 MicroVM의 AI 애플리케이션 수행 시 적합성을 평가하기 위한 성능 분석
Performance Analysis to Evaluate the Suitability of MicroVM with AI Applications for Edge Computing
0
조회
0
다운로드
0
인용
0
참고문헌
키워드
초록
본 논문에서는 엣지 컴퓨팅 환경에서 MicroVM의 AI 애플리케이션 수행 시 성능을 분석하고, 이것이 현재 사용되고 있는 컨테이너 기술과 전통적인 가상머신을 대체할 수 있는지 알아본다. 이를 위해 라즈베리파이 4에서 Docker 컨테이너, Firecracker MicroVM, KVM 가상머신 환경을 각각 구축하고 대표적인 AI 애플리케이션들을 실행하였다. 그리고 실험 환경별로 추론 시간, 총CPU 사용량 및 추세, 파일 I/O 성능을 분석하였다. 실험 결과, MicroVM에서 AI 애플리케이션을수행하였을 때 컨테이너와의 큰 성능 차이는 없었으며, 오히려 반복적인 애플리케이션 수행에서평균적으로 안정적인 추론 시간을 확인할 수 있었다. 따라서, 본 연구를 통해 엣지 컴퓨팅 환경에서 컨테이너와 가상머신을 대체하여 MicroVM을 사용한 AI 애플리케이션 운용이 적합할 수 있다는 것을 확인하였다.
Abstract
In this paper, we analyze the performance of MicroVM when running AI applications on an edge computing environment and whether it can replace current container technology and traditional virtual machines. To achieve this, we set up Docker container, Firecracker MicroVM and KVM virtual machine environments on a Raspberry Pi 4 and executed representative AI applications in each environment. We analyze the inference time, total CPU usage and trends over time and file I/O performance on each environment. The results show that there is no significant performance difference between MicroVM and container when running AI applications. Moreover, on average, a stable inference time over multiple trials was observed on MicroVM. Therefore, we can confirm that executing AI applications using MicroVM instead of container or heavy-weight virtual machine is suitable for an edge computing.
참고문헌 (21)
- J. A. Stankovic, "Research Directions for the Internet of Things", IEEE Internet of Things Journal, 1(1), pp.3~9 (2014)[DOI]
- "Number of Internet of Things (IoT) connected devices worldwide from 2019 to 2023, with forecasts from 2022 to 2030(in billions)"[URL]
- S. Al-Sarawi, "Internet of Things Market Analysis Forecasts, 2020–2030", Fourth World Conference on Smart Trends in Systems, Security and Sustainability (WorldS4), pp.449~453 (2020)
- A. Agache, "Firecracker : Lightweight virtualization for serverless applications", 17th USENIX NSDI, pp.419~434 (2020)
- "Kata Containers"[URL]
- M. Fezari, "Raspberry Pi 5 : The new Raspberry Pi family with more computation power and AI integration", DOI: 10.13140, 2023
- T. Hao, "Edge AIBench 2.0: A scalable autonomous vehicle benchmark for IoT–Edge–Cloud systems", BenchCouncil Transactions on Benchmarks, Standards and Evaluations, 2(4) (2022)[DOI]
- H. Hua, "Edge computing with artificial intelligence : A machine learning perspective", ACM Computing Surveys, 55(9), pp.1~35 (2023)[DOI]
- A. Villanueva, "Somnolence detection system utilizing deep neural network", International Conference on Information and Communications Technology (ICOIACT), pp.602~607 (2019), IEEE
- G. Lavanya, "An automated low cost IoT based fertilizer intimation system for smart agriculture", Sustainable Computing: Informatics and Systems, 28 (2020)[DOI]
- Anjali, T. Caraza-Harter, "Blending containers and virtual machines : a study of firecracker and gVisor", Proceedings of the 16th ACM SIGPLAN, 2020, 101~113
- Z. Wang, "Can"micro VM"become the next generation computing platform? : Performance comparison between light weight virtual machine, container, and traditional virtual machine", IEEE International Conference on Computer Science, Artificial Intelligence and Electronic Engineering (CSAIEE), pp.29~34 (2021)
- T. Goethals, "A functional and performance benchmark of lightweight virtualization platforms for edge computing", IEEE EDGE, pp.60~68 (2022)
- K. Lee, "MicroVM on Edge : Is It Ready for Prime Time?", International Symposium on the Modeling, Analysis, and Simulation of Computer and Telecommunication Systems (MASCOTS), pp.1~8 (2023)
- "Firecracker"[URL]
- J. Li, "DSFD : dual shot face detector", Proceedings of the IEEE, 2019, 5060~5069
- "DSFD-Pytorch-Inference"[URL]
- V. Jain, "Fddb : A benchmark for face detection in unconstrained settings", UMass Amherst technical report, 2(6), pp.1~11 (2010)
- "Silero Models: pre-trained enterprise-grade STT /TTS models and benchmarks"[URL]
- "The LJ Speech Dataset"[URL]