SHF: Small: Vertically Integrated Persistence

SHF：小型：垂直集成持久性

• 批准号: 1718335
• 负责人: Raju Rangaswami
• 金额: $ 45.8万
• 依托单位: Florida International University
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2017
• 资助国家: 美国
• 起止时间: 2017-09-15 至 2020-08-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1718335&HistoricalAwards=false
• 关键词: SHF; Small; Vertically; Integrated; Persistence

A new generation of data storage devices based on persistent memory technology are becoming available. Persistent memory devices address many shortcomings of previous generation technology including performance, power consumption, and device size. They have the potential to empower businesses and consumers with new and better capabilities. When these devices made available, either as part of a packaged digital service or within self-contained devices, they have the potential to vastly benefit the consumers of technology. This project builds the core software capabilities necessary for ensuring that the full potential of this new technology is realized and delivered to the end consumer. In particular, this project creates enabling technologies for operating data center servers, personal desktop and laptop computers, as well as mobile tablet computers and smart-phones more efficiently and at lower cost. Existing software support for storage are built around a set of core assumptions and goals that are rendered obsolete by persistent memory based storage devices. In this project, the PI will develop a new set of guiding principles for software that supports such devices from the ground up. In particular, this purpose-built software takes a vertically integrated approach using a minimalistic cross-layer design across components of the software stack. The new stack addresses the cornerstones of systems infrastructure including all levels of software and hardware, including application, compiler, the operating system, and the device access hardware infrastructure including the device itself. Educational activities include involvement of undergraduate students and incorporation of project research into courses taught by the PI. Outreach includes recruitment of under-represented minority students for participation in the project.

基于持久内存技术的新一代数据存储设备正在出现。持久内存设备解决了上一代技术的许多缺点，包括性能、功耗和设备尺寸。它们有潜力为企业和消费者提供新的、更好的功能。当这些设备可用时，无论是作为打包数字服务的一部分，还是作为独立设备的一部分，它们都有可能极大地造福技术消费者。该项目构建了核心软件功能，以确保实现这项新技术的全部潜力并将其交付给最终用户。特别是，该项目为数据中心服务器、个人台式机和笔记本电脑、移动平板电脑和智能手机的更高效、更低成本的操作创造了使能技术。对存储的现有软件支持是围绕一组核心假设和目标构建的，这些假设和目标已经被基于持久内存的存储设备淘汰了。在这个项目中，PI将开发一套新的指导原则，用于从头开始支持此类设备的软件。特别是，这个专门构建的软件采用垂直集成的方法，在软件堆栈的组件之间使用极简的跨层设计。新的堆栈解决了系统基础设施的基础，包括所有级别的软件和硬件，包括应用程序、编译器、操作系统和设备访问硬件基础设施，包括设备本身。教育活动包括本科生的参与，并将项目研究纳入PI教授的课程。外联活动包括招募代表性不足的少数民族学生参与该项目。