Summary
International Conference on Emerging Technologies for Communications
2021
Session Number:P4
Session:
Number:P4-22
AI-based Cache Replacement in Virtualized Environment
Taisei Miura, Kenichi Kourai, Saneyasu Yamaguchi ,
pp.-
Publication Date:2021/12/1
Online ISSN:2188-5079
Summary:
Deep learning has been producing significant breakthroughs in many fields, especially in image processing and natural language processing. However, it has been used in a restrained manner in a system software field. For example, deep learning may be suitably supported in cache replacement, but replacement using this has not been well studied. In usual cases, the least recently used (LRU) algorithm, which is traditionally and widely used, works very effectively. However, in the host operating system cache in a virtualized environment, LRU does not work well. In particular, the performance of LRU is severely low if the guest cache size is larger than the host cache size. Therefore, a proposal of a new algorithm, for example, an algorithm based on deep learning for such a situation is expected. In this paper, we focus on a cache replacement based on deep learning and discuss its performance.