Efficient implementation of interval matrix multiplication

Hong Diep Nguyen 1, *
* Corresponding author
1 ARENAIRE - Computer arithmetic
Inria Grenoble - Rhône-Alpes, LIP - Laboratoire de l'Informatique du Parallélisme
Abstract : The straightforward implementation of interval matrix product suf- fers from poor efficiency, far from the performances of highly optimized floating-point implementations. In this paper, we show how to reduce the interval matrix multiplication to 9 floating-point matrix products - for performance issues - without sacrificing the quality of the result. We show that, compared to the straightforward implementation, the overes- timation factor is at most 1.18.
Document type :
Conference papers
Complete list of metadatas

Cited literature [3 references]  Display  Hide  Download

https://hal.inria.fr/inria-00469472
Contributor : Hong Diep Nguyen <>
Submitted on : Thursday, April 1, 2010 - 3:09:55 PM
Last modification on : Thursday, February 7, 2019 - 2:24:28 PM
Long-term archiving on : Monday, July 5, 2010 - 9:02:49 PM

Files

NGUYEN_mat_prod.pdf
Files produced by the author(s)

Identifiers

  • HAL Id : inria-00469472, version 1

Collections

Citation

Hong Diep Nguyen. Efficient implementation of interval matrix multiplication. Para 2010: State of the Art in Scientific and Parallel Computing, Jun 2010, Reykjavik, Iceland. ⟨inria-00469472⟩

Share

Metrics

Record views

424

Files downloads

563