Collaborative Research: CISE: Large: Systems Support for Run-Anywhere Serverless

Computing resources are more abundant and more distributed than ever before. In principle, this abundance of compute and storage should allow applications to run faster, cheaper, and more energy efficiently. Yet applications that do so are few and far between, and usually belong to one of the very few large tech organizations. This project aims to democratize access to the full space of computing resources by enabling, and then optimizing, correct and safe execution of application logic anywhere: across client devices, edge computing locations, and multiple clouds. To realize the project’s execute-anywhere vision it will be necessary to develop a new co-design for four layers of system support: 1) a runtime that can execute applications anywhere; 2) an abundance of caches that enable seamless access to data anywhere; 3) a data store that collaborates with caches to ensure application consistency and data durability; and 4) a scheduler that integrates information from the other layers to decide where to run computations and move data in order optimize performance, availability, cost, and energy. The development of these co-designed layers will leverage the diverse expertise of the six investigators in close collaboration. Its development has the potential to improve the online applications that are an integral part of people’s lives. The project should make automatic optimization of an application’s latency, cost, bandwidth usage, and energy efficiency accessible to many applications, even those developed by a single person. The optimization of an application’s latency will improve the experience of all users with an especially large effect on users in rural communities and other locations with poor Internet connectivity. Further, it has the potential to make the full use and optimization of compute and data resources ubiquitous and available to all. In addition, it will advance discovery and broaden participation by involving undergraduates from the investigators’ universities and especially institutions that serve under-represented groups in research experiences, training graduate students, and broadening undergraduate participation in Computer Science. 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.