Improving Data Locality by Chunking - Inria - Institut national de recherche en sciences et technologies du numérique Access content directly
Conference Papers Year : 2003

Improving Data Locality by Chunking


Cache memories were invented to decouple fast processors from slow memories. However, this decoupling is only partial, and many researchers have attempted to improve cache use by program optimization. Potential benefits are significant since both energy dissipation and performance highly depend on the traffic between memory levels. But modeling the traffic is difficult; this observation has led to the use of heuristic methods for steering program transformations. In this paper, we propose another approach: we simplify the cache model and we organize the target program in such a way that an asymptotic evaluation of the memory traffic is possible. This information is used by our optimization algorithm in order to find the best reordering of the program operations, at least in an asymptotic sense. Our method optimizes both temporal and spatial locality. It can be applied to any static control program with arbitrary dependences. The optimizer has been partially implemented and applied to non-trivial programs. We present experimental evidence that the amount of cache misses is drastically reduced with corresponding performance improvements.
Fichier principal
Vignette du fichier
BastoulFeautrierCC03.pdf (203.94 Ko) Télécharger le fichier

Dates and versions

inria-00001055 , version 1 (24-01-2006)


  • HAL Id : inria-00001055 , version 1


Cédric Bastoul, Paul Feautrier. Improving Data Locality by Chunking. CC'2003 12th International Conference on Compiler Construction, Apr 2003, Warsaw/Poland. ⟨inria-00001055⟩
189 View
312 Download


Gmail Facebook Twitter LinkedIn More