A fully run-time auto-tuned sparse iterative solver with OpenATLib

Ken Naono, Takao Sakurai, Mitsuyoshi Igai, Takahiro Katagiri, Satoshi Ohshima, Shoji Itoh, Kengo Nakajima, Hisayasu Kuroda

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

We propose a general application programming interface called OpenATLib for auto-tuning (AT). OpenATLib is carefully designed to establish the reusability of AT functions for sparse iterative solvers. Using APIs of OpenATLib, we develop a fully auto-tuned sparse iterative solver called Xabclib. Xabclib has several novel runtime AT functions. We also develop a numerical computation policy that can optimize memory space and computational accuracy. Using the above functions and policies, we obtain the following important findings: (1) an average memory space is reduced to 1/45 under lower memory policies, and (2) fault convergence, which the conventional solvers judges to be converged but actually not converged in the sense of the before-preconditioned matrix, is avoided under higher accuracy policies. The results imply policy-based runtime AT plays significant role in sparse iterative matrix computations.

Original languageEnglish
Title of host publicationICIAS 2012 - 2012 4th International Conference on Intelligent and Advanced Systems
Subtitle of host publicationA Conference of World Engineering, Science and Technology Congress (ESTCON) - Conference Proceedings
Pages143-148
Number of pages6
DOIs
Publication statusPublished - Nov 1 2012
Event2012 4th International Conference on Intelligent and Advanced Systems, ICIAS 2012 - Kuala Lumpur, Malaysia
Duration: Jun 12 2012Jun 14 2012

Publication series

NameICIAS 2012 - 2012 4th International Conference on Intelligent and Advanced Systems: A Conference of World Engineering, Science and Technology Congress (ESTCON) - Conference Proceedings
Volume1

Other

Other2012 4th International Conference on Intelligent and Advanced Systems, ICIAS 2012
CountryMalaysia
CityKuala Lumpur
Period6/12/126/14/12

All Science Journal Classification (ASJC) codes

  • Artificial Intelligence
  • Information Systems

Fingerprint Dive into the research topics of 'A fully run-time auto-tuned sparse iterative solver with OpenATLib'. Together they form a unique fingerprint.

  • Cite this

    Naono, K., Sakurai, T., Igai, M., Katagiri, T., Ohshima, S., Itoh, S., Nakajima, K., & Kuroda, H. (2012). A fully run-time auto-tuned sparse iterative solver with OpenATLib. In ICIAS 2012 - 2012 4th International Conference on Intelligent and Advanced Systems: A Conference of World Engineering, Science and Technology Congress (ESTCON) - Conference Proceedings (pp. 143-148). [6306176] (ICIAS 2012 - 2012 4th International Conference on Intelligent and Advanced Systems: A Conference of World Engineering, Science and Technology Congress (ESTCON) - Conference Proceedings; Vol. 1). https://doi.org/10.1109/ICIAS.2012.6306176