A National-Scale Testbed Supporting Artificial Intelligence Research Spanning the Computing Continuum

支持跨越计算连续体的人工智能研究的国家级测试平台

• 批准号: 2331263
• 负责人: Peter Beckman
• 金额: $ 320万
• 依托单位: Northwestern University
• 依托单位国家: 美国
• 项目类别: Continuing Grant
• 财政年份: 2023
• 资助国家: 美国
• 起止时间: 2023-09-15 至 2026-07-31
• 项目状态: 未结题
• 来源: https://www.nsf.gov/awardsearch/showAward?AWD_ID=2331263&HistoricalAwards=false
• 关键词: National; Scale; Testbed; Supporting; Artificial

New artificial intelligence (AI) is rapidly transforming how the nation conducts scientific research. However, as scientists quickly adopt new AI techniques, the architecture and design of a performant computational infrastructure to support these new explorations, particularly as NSF has rapidly grown a new community of AI research teams through programs such as the AI institutes, represent fundamental open questions. Developing and evaluating new approaches to AI-enabled scientific workflows, programming and runtime environments, integrating data from diverse sources–in context of a national-scale computing continuum spanning intelligent edge devices to centralized high-performance computing centers–require at-scale test Cyberinfrastructure.This project designs and builds a new national-scale testbed that supports computer scientists developing and exploring novel distributed-AI computing systems. The edge-based pilot testbed links low-power AI-enabled computers distributed across the country with high-resolution environmental sensors and instruments such as cameras, microphones, and weather sensors. The testbed supports scientists exploring design concepts, operational modes, and real-time sensor data workflows, contributing to the understanding of Earth's changing climate and biosphere. The pilot testbed also addresses today’s urgent need for a diverse, AI-trained scientific workforce by enabling students, researchers, and community scientists to ethically engage in multidisciplinary AI research, creating the next generation of scientists needed to responsibly develop AI capabilities in every facet of our economy. By addressing the current lack of a comprehensive, flexible toolset and infrastructure for AI-centric and computationally mediated research, the project provides a crucial research testbed that bridges current computational resources and the demands of advancing AI capabilities. This effort is also supported by National Discovery Cloud for Climate (NDC-C) resources.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.

新的人工智能（AI）正在迅速改变国家进行科学研究的方式。 然而，随着科学家们迅速采用新的人工智能技术，支持这些新探索的高性能计算基础设施的架构和设计，特别是NSF通过人工智能研究所等项目迅速发展了一个新的人工智能研究团队社区，代表了基本的开放问题。 开发和评估支持人工智能的科学工作流程、编程和运行时环境的新方法，整合来自不同来源的数据--在从智能边缘设备到集中式高性能计算中心的国家级计算连续体的背景下--需要大规模测试Cyberinfrastructure。该项目设计并构建了一个新的国家级测试平台，支持计算机科学家开发和探索新型分布式人工智能计算系统。 基于边缘的试点测试平台将分布在全国各地的低功耗人工智能计算机与高分辨率环境传感器和仪器（如相机，麦克风和天气传感器）连接起来。 该试验台支持科学家探索设计概念、操作模式和实时传感器数据工作流程，有助于了解地球不断变化的气候和生物圈。该试点测试平台还解决了当今对多样化、受过人工智能培训的科学劳动力的迫切需求，使学生、研究人员和社区科学家能够在道德上参与多学科人工智能研究，培养下一代科学家，以负责任的方式在我们经济的各个方面发展人工智能能力。通过解决目前缺乏以人工智能为中心和以计算为中介的研究的全面，灵活的工具集和基础设施的问题，该项目提供了一个关键的研究测试平台，可以连接当前的计算资源和提高人工智能能力的需求。这项工作也得到了国家探索云气候（NDC-C）资源的支持。该奖项反映了NSF的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。