This Project Grant award from the National Science Foundation Division of Computing and Communication Foundations provides $125,000 in funding to Oregon State University from October 1, 2022 to September 30, 2023. The award supports research into cross-layer coordination and optimization for scalable and sparse tensor networks under the Computer and Information Science and Engineering federal grant program (CFDA 47.070).
Specifically, the award will fund preliminary research exploring memory heterogeneity-aware representations and data rearrangement, balanced sparse tensor contraction algorithms with optimized data placement, memoization techniques to reduce computational costs, and specialized accelerator architectures for sparse tensor networks. The optimized approaches are intended to enable faster and more memory-efficient computation of sparse tensor networks, which are increasingly important for high-dimensional data analysis but pose computational challenges. Insights from this planning project could help advance tensor decomposition methods and trigger further collaboration across academia, research laboratories, and industry.