Corpus ID: 209531592

A Unified Iteration Space Transformation Framework for Sparse and Dense Tensor Algebra

@article{Senanayake2020AUI,
  title={A Unified Iteration Space Transformation Framework for Sparse and Dense Tensor Algebra},
  author={Ryan Senanayake and Fredrik Kjolstad and ChangWan Hong and Shoaib Kamil and Saman P. Amarasinghe},
  journal={ArXiv},
  year={2020},
  volume={abs/2001.00532}
}
  • Ryan Senanayake, Fredrik Kjolstad, +2 authors Saman P. Amarasinghe
  • Published 2020
  • Computer Science
  • ArXiv
  • We address the problem of optimizing mixed sparse and dense tensor algebra in a compiler. We show that standard loop transformations, such as strip-mining, tiling, collapsing, parallelization and vectorization, can be applied to irregular loops over sparse iteration spaces. We also show how these transformations can be applied to the contiguous value arrays of sparse tensor data structures, which we call their position space, to unlock load-balanced tiling and parallelism. We have prototyped… CONTINUE READING

    Citations

    Publications citing this paper.

    References

    Publications referenced by this paper.
    SHOWING 1-10 OF 64 REFERENCES

    Tensor Algebra Compilation with Workspaces

    VIEW 4 EXCERPTS

    The tensor algebra compiler

    VIEW 6 EXCERPTS

    SPLATT: Efficient and Parallel Sparse Tensor-Matrix Multiplication

    VIEW 1 EXCERPT

    Optimizing Sparse Matrix Computations for Register Reuse in SPARSITY

    VIEW 1 EXCERPT

    Implementing sparse matrix-vector multiplication on throughput-oriented processors

    • Nathan Bell, Michael Garland
    • Computer Science
    • Proceedings of the Conference on High Performance Computing Networking, Storage and Analysis
    • 2009
    VIEW 5 EXCERPTS
    HIGHLY INFLUENTIAL

    Merge-Based Parallel Sparse Matrix-Vector Multiplication

    • Duane Merrill, Michael Garland
    • Computer Science
    • SC16: International Conference for High Performance Computing, Networking, Storage and Analysis
    • 2016
    VIEW 6 EXCERPTS
    HIGHLY INFLUENTIAL

    Load-Balanced Sparse MTTKRP on GPUs

    VIEW 18 EXCERPTS
    HIGHLY INFLUENTIAL

    Automating Wavefront Parallelization for Sparse Matrix Computations

    VIEW 1 EXCERPT