NSF News

NSF invests $36M in computing projects that promise to maximize performance, reduce energy demands

3 projects receive awards through the Expeditions in Computing program

The U.S. National Science Foundation is awarding $36 million to three projects selected for their potential to revolutionize computing and make significant impacts in reducing the carbon footprint of the lifecycle of computers. Funding for the projects comes from the NSF Expeditions in Computing (Expeditions) program, an ambitious initiative that supports transformative research poised to yield lasting impacts on society, the economy and technological advancement. Projects funded by Expeditions are characterized by their ambition and potential for transformation, leveraging advances in computing and cyberinfrastructure to accelerate discovery and innovation across various domains of science and engineering. 

"We are thrilled to announce these visionary projects that will advance environmental responsibility and foster innovation in the field of computing," said Dilma DaSilva, acting assistant director for the NSF Directorate for Computer and Information Science and Engineering (CISE). "Congratulations to these pioneering teams whose research will forge new pathways in computational decarbonization and in revolutionizing operating system design with machine learning.  

The 2024 NSF Expeditions awardees  

NSF Expeditions in Computing: Carbon Connect--An Ecosystem for Sustainable Computing. Led by Harvard University, this multi-institutional, five-year research initiative will lay the foundations for sustainable computing, with a focus on reducing the environmental impact of computer systems. This shift toward sustainability could spark a transformation in how computer systems are manufactured, allocated and consumed, leading to a more responsible and sustainable approach to advancing computing technologies. By redefining the way computer scientists consider environmental sustainability, Carbon Connect will establish new standards for carbon accounting in the computing industry, thereby influencing future energy policy and legislation. 

Collaborators on this project include the University of Pennsylvania, the California Institute of Technology, Carnegie Mellon University, Cornell University, Yale University and The Ohio State University. 

NSF Expeditions in Computing for Computational Decarbonization of Societal Infrastructures at Mesoscales. Led by the University of Massachusetts Amherst, this project will develop the new field of computational decarbonization, (CoDec), which focuses on optimizing and reducing the lifecycle of carbon emissions of complex computing and societal infrastructure systems. CoDec will tackle interdependencies across multiple aspects of infrastructure, including computing, transportation, buildings and the electric power grid. Through innovative sensing approaches, optimization methods grounded in theory and artificial intelligence, and software-defined interfaces, CoDec seeks to automate and coordinate carbon-efficiency optimizations across time, space and sectors. These efforts will enable scientific discoveries in decarbonization while supporting sustainable growth, advancing technology and strengthening national security.   

Collaborators on this project include Carnegie Mellon University, the Massachusetts Institute of Technology, the University of Chicago, UCLA and the University of Wisconsin-Madison. 

NSF Expeditions in Computing: Learning Directed Operating System--A Clean-Slate Paradigm for Operating Systems Design and Implementation. Led by The University of Texas at Austin, this project aims to revolutionize the design of operating systems (OSes) by integrating advanced machine learning (ML) into resource management. Current OSes employ rigid, manually designed approaches for allocating hardware resources among running applications. This inflexibility makes it hard to adapt to evolving application needs and hardware, leading to inefficiency and poor performance. This research will develop a learning-directed operating system with intrinsic intelligence and auto-adaptation, enabling ML-driven resource management that optimizes performance and efficiency and requires minimal human intervention. By fundamentally rethinking OS design with ML at its core, this research has the potential to significantly improve the energy efficiency of cloud computing, enable real-time edge computing applications and create innovative computing devices. 

Collaborators on this project include the University of Illinois Urbana-Champaign, the University of Pennsylvania, and University of Wisconsin-Madison.

About the Expeditions program 

Established in 2008, the NSF Expeditions awards represent some of the largest investments provided by the CISE directorate. Pioneering work funded by the program includes the Robobee Project and CompSustNet.  

Expeditions projects focus on creating transformative technologies, methodologies and infrastructure that can be adopted by the broader research community, industry or society at large. The program emphasizes the translation of research outcomes into practical applications, thus driving advancements in computer science and its real-world applications.