SHF: Small: High-Performance, Energy-Efficient Mobile Web Computing

SHF：小型：高性能、高能效的移动网络计算

• 批准号: 1619283
• 负责人: Vijay Janapa Reddi
• 金额: $ 40万
• 依托单位: University of Texas at Austin
• 依托单位国家: 美国
• 项目类别: Standard Grant
• 财政年份: 2016
• 资助国家: 美国
• 起止时间: 2016-06-15 至 2019-05-31
• 项目状态: 已结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=1619283&HistoricalAwards=false
• 关键词: SHF; Small; Performance; Energy; Efficient

The Web is continuously transforming society - shaping communications, catalyzing innovations, even shaping thought processes. Over the past decade, the role of the Web has shifted from information retrieval (Web 1.0) to providing a platform for interactive and engaging user experiences (Web 2.0). The Web is once again entering a new age, transcending user engagement to provide intelligent services that integrate multiple devices together. The driving force behind the Web's evolution is the ubiquity of mobile devices-undoubtedly today's most pervasive personal computing platform. The PI will study the crucial conjunction between mobile devices and Web technologies because mobile devices struggle to deliver high computational capability on a battery-power budget. The PI's integrated education plan, involving workshops, new curriculum and cutting-edge Web technologies, will stimulate students from a young age, who come from a diverse range of backgrounds, ethnicities, age groups and skill sets, to pursue academic interests in relevant scientific and technological fields. The exposure will nurture students to embrace careers in science, technology, engineering, and mathematics. The research will lay the groundwork and establish the foundations needed to develop future high-performance and energy-efficient mobile computing systems that can usher in the next generation of Web applications.  The PI will research and investigate optimizations across the layers of the execution stack: at the (1) hardware layer the research will yield insights into new mobile processor architectures that achieve high efficiency by exploiting domain-specific and runtime knowledge; at the (2) runtime software layer, the project will lay the foundations for feedback-driven mechanisms that can dynamically learn to operate the architecture at energy-efficient points without compromising end-user application performance and user experience; at the (3) application layer, the research will discover language hints to empower application writers with awareness of their applications' energy consumption.

网络正在不断改变社会 - 塑造交流，催化创新，甚至塑造思维过程。在过去的十年中，Web的角色已从信息检索（Web 1.0）转变为为交互式和引人入胜的用户体验提供平台（Web 2.0）。网络再次进入新时代，超越用户参与度，以提供将多个设备集成在一起的智能服务。网络演变背后的推动力是移动设备的普遍存在 - 毫无疑问，当今最普遍的个人计算平台。 PI将研究移动设备和Web技术之间的关键连接，因为移动设备难以在电池功率预算上提供高计算能力。 PI的综合教育计划涉及研讨会，新课程和尖端的Web技术，将刺激来自年轻的学生，他们来自潜水员的背景，种族，年龄段和技能，以追求相关科学和技术领域的学术利益。曝光将为学生提供科学，技术，工程和数学的职业。这项研究将奠定基础，并建立开发未来的高性能和节能移动计算系统所需的基础，这些计算系统可以引入下一代Web应用程序。 PI将研究和调查执行堆栈层面的优化：在（1）硬件层上，研究将通过利用特定领域和运行时知识来实现​​新的移动处理器体系结构的见解；在（2）运行时软件层，该项目将为反馈驱动的机制奠定基础，这些机制可以动态地学习以节能点操作体系结构，而不会损害最终用户应用程序性能和用户体验；在（3）应用层，该研究将发现语言提示，以使应用程序作者能够意识到其应用程序的能耗。