RVO: Unleashing SSD's Parallelism by Harnessing the Unused Power

Hasan Alhasan, Yun Chih Chen, Chien Chung Ho

研究成果: Conference contribution

摘要

Analytic video surveillance system is one of the fastest-growing cyber-physical applications worldwide. A video surveillance system must have a scalable storage backend to simultaneously ingest video frames and serve read requests for video analytics. As a result, 3D NAND flash-based storage devices, i.e., Solid-State Drives (SSD), are gradually regarded as promising candidates thanks to their rapidly growing density and parallelism. However, when deployed in power-constrained environments like the network edges, SSDs' parallelism often cannot be fully unleashed. In edge video analytics systems, the lost parallelism can cause video frame drop and untimely analytics. To tackle this limitation, we first reveal that a flash program operation's power usage is over-estimated in the conventional SSD design, leading to a limited degree of I/O parallelism. Based on the observation, we propose a novel command set, RVO (Read-Verify Overlap), which reclaims the unused power from the overestimation to amend the lost parallelism. To realize feasible fine-grained power management, we further accompany RVO with a generic power-aware scheduler. Through experiments, we show how video analytics systems equipped with RVO can achieve zero frame drop while ensuring compliance with industrial read latency requirements, even in write-intensive workloads.

原文English
主出版物標題2021 IEEE/ACM International Symposium on Low Power Electronics and Design, ISLPED 2021
發行者Institute of Electrical and Electronics Engineers Inc.
ISBN(電子)9781665439220
DOIs
出版狀態Published - 2021 7月 26
事件2021 IEEE/ACM International Symposium on Low Power Electronics and Design, ISLPED 2021 - Boston, United States
持續時間: 2021 7月 262021 7月 28

出版系列

名字Proceedings of the International Symposium on Low Power Electronics and Design
2021-July
ISSN(列印)1533-4678

Conference

Conference2021 IEEE/ACM International Symposium on Low Power Electronics and Design, ISLPED 2021
國家/地區United States
城市Boston
期間21-07-2621-07-28

All Science Journal Classification (ASJC) codes

  • 工程 (全部)

指紋

深入研究「RVO: Unleashing SSD's Parallelism by Harnessing the Unused Power」主題。共同形成了獨特的指紋。

引用此