We introduce Stardust, a compiler that compiles sparse tensor algebra to reconfigurable dataflow architectures (RDAs). Stardust introduces new user-provided data representation and scheduling language constructs for mapping to resource-constrained accelerated architectures. Stardust uses the information provided by these constructs to determine on-chip memory placement and to lower to the Capstan RDA through a parallel-patterns rewrite system that targets the Spatial programming model. The Stardust compiler is implemented as a new compilation path inside the TACO open-source system. Using cycle-accurate simulation, we demonstrate that Stardust can generate more Capstan tensor operations than its authors had implemented and that it results in 138× better performance than generated CPU kernels and 41× better performance than generated GPU kernels.
@article{arxiv.2211.03251,
title = {Stardust: Compiling Sparse Tensor Algebra to a Reconfigurable Dataflow Architecture},
author = {Olivia Hsu and Alexander Rucker and Tian Zhao and Kunle Olukotun and Fredrik Kjolstad},
journal= {arXiv preprint arXiv:2211.03251},
year = {2022}
}