Learning to Drive Software-Defined Solid-State Drives
Learning to Drive Software-Defined Solid-State Drives
复制标题
学习驱动软件定义的固态硬盘
DOI：
10.1145/3613424.3614281
复制
发表时间：
2023
期刊：
影响因子：
--
通讯作者：
Huang, Jian
中科院分区：
文献类型：
--
作者：
Li, Daixuan;Sun, Jinghan;Huang, Jian
得益于成熟的制造技术，基于闪存的固态硬盘（SSD）可针对当今的应用进行高度定制，这为进一步提高其存储性能和资源利用率带来了机会。然而，SSD的效率通常是由许多硬件参数，使开发人员很难手动调整它们，并确定优化的SSD硬件configurations.In本文中，我们提出了一个自动学习为基础的SSD硬件配置框架，名为AutoBlox，利用监督和无监督机器学习（ML）技术来驱动SSD的硬件配置的调整。AutoBlox使用其数据块I/O跟踪自动提取新工作负载的独特访问模式，将工作负载映射到以前的工作负载以利用学习到的经验，并根据验证的存储性能推荐优化的SSD配置。AutoBlox通过自动化硬件参数配置和减少手动工作来加速新SSD设备的开发。我们使用简单而有效的学习算法开发AutoBlox，这些算法可以在多核CPU上高效运行。在给定目标存储工作负载的情况下，我们的评估表明，AutoBlox可以提供优化的SSD配置，与商用SSD相比，可以将目标工作负载的性能平均提高1.30倍，同时满足SSD容量、设备接口和功耗预算等特定约束。此配置将最大限度地提高目标工作负载和非目标工作负载的性能。
Thanks to the mature manufacturing techniques, flash-based solid-state drives (SSDs) are highly customizable for applications today, which brings opportunities to further improve their storage performance and resource utilization. However, the SSD efficiency is usually determined by many hardware parameters, making it hard for developers to manually tune them and determine the optimized SSD hardware configurations.In this paper, we present an automated learning-based SSD hardware configuration framework, named AutoBlox, that utilizes both supervised and unsupervised machine learning (ML) techniques to drive the tuning of hardware configurations for SSDs. AutoBlox automatically extracts the unique access patterns of a new workload using its block I/O traces, maps the workload to previous workloads for utilizing the learned experiences, and recommends an optimized SSD configuration based on the validated storage performance. AutoBlox accelerates the development of new SSD devices by automating the hardware parameter configurations and reducing the manual efforts. We develop AutoBlox with simple yet effective learning algorithms that can run efficiently on multi-core CPUs. Given a target storage workload, our evaluation shows that AutoBlox can deliver an optimized SSD configuration that can improve the performance of the target workload by 1.30 × on average, compared to commodity SSDs, while satisfying specified constraints such as SSD capacity, device interfaces, and power budget. And this configuration will maximize the performance improvement for both target workloads and non-target workloads.
登录
查看更多内容
影响因子：
2
作者：
Sylvain Firer-Blaess;C. Fuchs
通讯作者：
Sylvain Firer-Blaess;C. Fuchs
DOI：
10.1109/micro.2018.00045
发表时间：
2018-10
期刊：
2018 51st Annual IEEE/ACM International Symposium on Microarchitecture (MICRO)
影响因子：
--
作者：
Donghyun Gouk;Miryeong Kwon;Jie Zhang;Sungjoon Koh;Wonil Choi;N. Kim;M. Kandemir;Myoungsoo Jung
通讯作者：
Donghyun Gouk;Miryeong Kwon;Jie Zhang;Sungjoon Koh;Wonil Choi;N. Kim;M. Kandemir;Myoungsoo Jung
DOI：
10.1145/3302424.3303983
发表时间：
2019-03
期刊：
Proceedings of the Fourteenth EuroSys Conference 2019
影响因子：
--
作者：
Xiaohao Wang;Yifan Yuan;You Zhou;Chance C. Coats;Jian Huang
通讯作者：
Xiaohao Wang;Yifan Yuan;You Zhou;Chance C. Coats;Jian Huang
设计是……
DOI：
--
发表时间：
2023
期刊：
The Design Journal
影响因子：
--
作者：
Louise Valentine
通讯作者：
Louise Valentine