ASaP: Automatic Software Prefetching for Sparse Tensor Computations in MLIR
Sparse tensor computations suffer from irregular memory access patterns that degrade cache performance. While software prefetching can mitigate this, existing compiler approaches lack the semantic insight needed for effective optimization.
Authors
- Jonas Skeppstedt (ORCID: https://orcid.org/0009-0001-4549-6260)
- Per Stenström (ORCID: https://orcid.org/0000-0002-7441-8245)
- Konstantinos Sotiropoulos (ORCID: https://orcid.org/0000-0002-6603-1171)
Institutions
- Lund University (SE)
- Chalmers University of Technology (SE)
- University of Gothenburg (SE)
Publication Details
- Published
- 2025-11-07
- DOI
- https://doi.org/10.1145/3731599.3767477
- Citations
- 1
- Primary Topic
- Parallel Computing and Optimization Techniques
- Type
- article
- Field-Weighted Citation Impact
- 2.27