Collaborative Research: CSR: Medium: Towards A Unified Memory-centric Computing System with Cross-layer Support

Data-centric applications, from scientific simulation to emerging machine learning and data mining algorithms, are becoming prevalent. Memory-centric computing is a potential solution to overcome the unprecedented memory performance bottleneck for such applications. We propose an integrated, full-stack, cross-layer system to enable Unified Memory-centric Computing (UniMCC). We target systems that have near-memory data processors (NDPs) as well as a disaggregated shared memory pool. We work on the entire system stack to utilize these newly proposed hardware solutions from a unified viewpoint of architecture, SW/HW interface, code generation and runtime support, and performance modeling and optimization. The goal of the proposed UniMCC system is to maximize the potential of NDPs and disaggregated memories and lift memory-centric computing to a new level. The outcomes of the project will have significant broader impacts on research communities, society, and education. With the cross-layer collaborative design, the proposed research can bring state-of-the-art techniques together to make the memory-centric paradigm more feasible. The success of the project will enable experiments and deployment of data-centric applications at larger scales, facilitating research in the AI area and improving the availability of large-scale data mining and machine learning systems for solving real-world problems. Besides training Ph.D. students, this project will deliver several course developments, curricular update activities, and outreach activities to broaden participation in computing. 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.