# Cache oblivious storage and access heuristics for blocked matrix-matrix multiplication

Preprint · Nicolas Bock, Emanuel H. Rubensson, Paweł Sałek, Anders M. N. Niklasson, Matt Challacombe

- Archive id: arxiv:0808.1108
- Published: 2008-08-07
- arXiv: https://arxiv.org/abs/0808.1108
- Source URL: https://arxiv.org/abs/0808.1108
- PDF: https://arxiv.org/pdf/0808.1108
- Provenance: arxiv

## Abstract

We investigate effects of ordering in blocked matrix--matrix multiplication. We find that submatrices do not have to be stored contiguously in memory to achieve near optimal performance. Instead it is the choice of execution order of the submatrix multiplications that leads to a speedup of up to four times for small block sizes. This is in contrast to results for single matrix elements showing that contiguous memory allocation quickly becomes irrelevant as the blocksize increases.

_Abstract reproduced under CC0-1.0 (arXiv descriptive metadata)._

---
Space Kitty Archive · https://space-kitty.com · API: https://space-kitty.com/api/v1 · MCP: https://space-kitty.com/mcp · Premium: $0.20/request via x402 (Base Sepolia (testnet)). Metadata: arXiv (CC0), OpenAlex (CC0), Europe PMC (CC BY/CC0 works), ClinicalTrials.gov.