Frontiers in Educational Research, 2022, 5(12); doi: 10.25236/FER.2022.051216.
Chunlin Chen
Economic Management Experimental Teaching Center, Southwest University of Finance and Economics, Chengdu, 611130, China
In the process of integration of AI teaching into various disciplines, theoretical teaching and practical teaching are given equal importance, while the experimental teaching of AI has specific and complex requirements for the experimental environment. This paper introduces the practice and reflection on the method of microservice cluster transformation into AI experimental teaching platform based on the traditional laboratory, which makes full use of the existing laboratory hardware resources, fully introduces K8s, Docker and other open source technology solutions, creates a full process experimental environment for AI teaching that can fully improve the efficiency of teaching and research without affecting the daily teaching activities, optimizes the operation mechanism and service mode, realize the on-demand customization of teaching resources, improve teaching governance, and provide useful ideas and methods for the construction of experimental teaching environment of artificial intelligence.
artificial intelligence, experimental teaching, teaching platform, microservices, K8s, Jupyter, vGPU
Chunlin Chen. A Practical Study on Building a Distributed Artificial Intelligence Experimental Teaching Platform Based on Traditional Laboratories. Frontiers in Educational Research (2022) Vol. 5, Issue 12: 87-91. https://doi.org/10.25236/FER.2022.051216.
[1] N. Zeng, Y. Zhang. (2022). Ensuring AI serves the common good and promotes education system change-"Artificial intelligence and education: a guide for policymakers" interpreted. China's E-learning, no.427, p.1-8.
[2] J. X. Wang, L. Guo. (2018). Enterprise container cloud platform solution based on Kubernetes and Docker technology. Journal of Shanghai Institute of Shipping and Transportation Science, no.41(3), p. 51-57.
[3] Z. X. Wu. (2017). Development and trends of cloud computing virtualization technology. Computer Applications, no. 37(4), p. 915-923.
[4] J.W. Miao, Z. Wang, J. Yu et al. (2019). Performance analysis of high-performance vGPU computing in desktop cloud environment. Computer Technology and Development, no. 29(11) p. 184-189.
[5] W. N. Li. (2018). Applying differential update techniques to achieve performance optimization of GPU virtualization scenarios. Shanghai Jiao Tong University.
[6] X. Ying. (2013). How to build and configure NFS (Network File System) under Linux. Computers and Networks, no. 39(21) p.44-45.
[7] Y.X. Lu. (2021). Machine learning platform improvements for distributed training and high-performance inference with Tensorflow. Beijing Jiaotong University.
[8] Y.P. Huang, W.X. Liang, Z.H. Xiao. (2020). A comparative analysis of deep learning frameworks based on TensorFlow and PyTorch. Modern Information Technology, no. 4(4). p. 80-82,87.
[9] Z. P. He, X. D. Zhang, Y. Liu. (2019). Microservice architecture based on Jupyter interactive analysis platform. Computer System Applications. no. 28(8) p.63-70.
[10] Y. Q. Zhu, X. Tang. (2016). Analysis and research on VNC-based remote desktop transfer protocol. Computer System Applications, no. 25(11). p. 284-287.
[11] F.H. Liu,H.L. Ding, Y.P. Zhang. (2006). Research and application on NAT technology. Computer Engineering and Design, no.10. p.1814-1817.
[12] N. Fei, H.R. Zhang. (2019). A study of TensorFlow architecture and implementation mechanism. Computer Technology and Development, no.29(9). p. 31-34.
[13] L. Ma, J. D. Song, M. N. Song. (2012). Open platform:A review of operation model and technical architecture research. Telecommunications Science, no.28(6), p.125-140.