Skip to Main content Skip to Navigation

High Performance by Exploiting Information Locality through Reverse Computing

Mouad Bahi 1 Christine Eisenbeis 1 
1 ALCHEMY - Architectures, Languages and Compilers to Harness the End of Moore Years
LRI - Laboratoire de Recherche en Informatique, UP11 - Université Paris-Sud - Paris 11, Inria Saclay - Ile de France, CNRS - Centre National de la Recherche Scientifique : UMR8623
Abstract : In this paper we present performance results for our register rematerialization technique based on reverse recomputing. Rematerialization adds instructions and we show on one specifically designed example that reverse computing alleviates the impact of these additional instructions on performance. We also show how thread parallelism may be optimized on GPUs by performing register allocation with reverse recomputing that increases the number of threads per Streaming Multiprocessor (SM). This is done on the main kernel of Lattice Quantum ChromoDynamics (LQCD) simulation program where we gain a 10.84% speedup.
Complete list of metadata

Cited literature [10 references]  Display  Hide  Download
Contributor : Mouad Bahi Connect in order to contact the contributor
Submitted on : Friday, August 19, 2011 - 3:53:35 PM
Last modification on : Sunday, June 26, 2022 - 11:54:13 AM
Long-term archiving on: : Monday, November 12, 2012 - 3:36:21 PM


Files produced by the author(s)


  • HAL Id : inria-00615493, version 1



Mouad Bahi, Christine Eisenbeis. High Performance by Exploiting Information Locality through Reverse Computing. [Research Report] 2011. ⟨inria-00615493⟩



Record views


Files downloads