Mosaic: An Interoperable Compiler for Tensor Algebra

Manya Bansal; Olivia Hsu; Kunle Olukotun; Fredrik Kjolstad

doi:10.1145/3591236

PLDI · June 2023

Mosaic: An Interoperable Compiler for Tensor Algebra

Manya Bansal Olivia Hsu Kunle Olukotun Fredrik Kjolstad

ACM SIGPLAN Conference on Programming Language Design and Implementation · PACMPL vol. 7, PLDI

Distinguished Paper Award

Read PDF DOI

Topics Sparse tensor algebra Compiler representations

Related software

Mosaic: An Interoperable Compiler for Tensor Algebra A tensor algebra compiler, built on TACO, that binds parts of an expression to external hand-tuned functions from other libraries and generates the rest, searching for the best mapping.
TACO: The Tensor Algebra Compiler A compiler that generates fast kernels for tensor algebra expressions over dense and sparse tensors from a high-level index notation.

Cite this paper

BibTeX

@article{bansal2023mosaic,
  author    = {Bansal, Manya and Hsu, Olivia and Olukotun, Kunle and Kjolstad, Fredrik},
  title     = {{Mosaic}: An Interoperable Compiler for Tensor Algebra},
  journal   = {Proceedings of the ACM on Programming Languages},
  year      = {2023},
  month     = jun,
  volume    = {7},
  number    = {PLDI},
  articleno = {122},
  numpages  = {26},
  publisher = {Association for Computing Machinery},
  address   = {New York, NY, USA},
  doi       = {10.1145/3591236},
  url       = {https://doi.org/10.1145/3591236}
}

Citation

Manya Bansal, Olivia Hsu, Kunle Olukotun, and Fredrik Kjolstad. Mosaic: An Interoperable Compiler for Tensor Algebra. PLDI, 2023.

All publications