CAREER: Enabling Memory-Centric Computing from Internet of Things to Cloud

职业：实现从物联网到云的以内存为中心的计算

• 批准号: 2339317
• 负责人: Elaheh Sadredini
• 金额: $ 58.76万
• 依托单位: University of California-Riverside
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2024
• 资助国家: 美国
• 起止时间: 2024-04-01 至 2029-03-31
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2339317&HistoricalAwards=false
• 关键词: CAREER; Enabling; Memory; Centric; Computing

Today's computing landscape is predominantly driven by data-centric applications, where their performance heavily relies on efficient data access and processing. Memory-centric computing revolutionizes data processing by performing computations in close proximity to the locations where data is generated/captured or stored, such as sensors, memory, and storage devices, and capitalizes on the extensive parallelism inherent in memory arrays. Therefore, it fundamentally tackles the data movement and instruction processing bottleneck, offering a powerful solution for data-intensive workloads in domains like machine learning and genomics, with far-reaching implications for scientific advancements. The overarching goal of this project is to facilitate the paradigm shift from compute-centric to memory-centric systems and seamlessly and securely integrate them into today's modern systems. This transformation will enable efficient and secure computation, fostering advancements in data processing capabilities. This project aims to integrate the acquired methodologies and insights into age-appropriate curricula, from K-12 to graduate-level courses. It also incorporates several STEM (science, technology, engineering and mathematics) enrichment activities for local K-12 and community college students in the Inland Empire.This award aims to facilitate the integration of the memory-centric computing paradigm into present-day computing systems. To fully exploit the potential of memory-centric computing, this project takes a holistic approach by co-designing the entire computing stack to facilitate the adoption of near-data processing in modern architectures. It will also investigate the security and privacy challenges posed by memory-centric computing solutions and propose lightweight techniques to effectively address these challenges. The developed frameworks and tools will be made publicly available to benefit the computer science community and to serve as the foundation for new course development and creating research opportunities for undergraduate and high school students.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

当今的计算环境主要是由以数据为中心的应用程序驱动的，它们的性能在很大程度上依赖于有效的数据访问和处理。以内存为中心的计算通过密切靠近数据生成/捕获或存储数据的位置（例如传感器，内存和存储设备）并在内存阵列中固有的广泛的并行性来大写，从而彻底改变了数据处理。因此，它从根本上解决了数据运动和指导处理瓶颈，为机器学习和基因组学等领域的数据密集型工作量提供了强大的解决方案，对科学进步具有深远的影响。该项目的总体目标是促进范式从以计算为中心到以内存为中心的系统的范式转移，并将它们无缝地整合到当今的现代系统中。这种转换将实现高效且安全的计算，从而促进数据处理能力的进步。该项目旨在将获得的方法和见解纳入适合年龄的课程，从K-12到研究生级课程。它还为本地帝国的本地K-12和社区大学生提供了几种STEM（科学，技术，工程和数学）丰富的活动。该奖项旨在促进以内存为中心的计算范式整合到当今的计算系统中。为了充分利用以内存为中心的计算的潜力，该项目通过共同设计整个计算堆栈来采用整体方法，以促进在现代体系结构中采用近DATA处理。它还将调查以内存计算解决方案提出的安全性和隐私挑战，并提出轻巧的技术以有效解决这些挑战。开发的框架和工具将被公开使用，以使计算机科学界有利，并为新课程开发的基础，并为本科和高中生创造研究机会。该奖项反映了NSF的法定任务，并被认为是通过基金会的知识分子优点和更广泛的审查标准来通过评估来获得支持的。