The Community for Technology Leaders
RSS Icon
Subscribe
Issue No.01 - January-June (2011 vol.10)
pp: 20-23
Chunyang Gou , Delft University of Technology, Delft
Georgi Gaydadjiev , Delft University of Technology, Delft
ABSTRACT
In this paper, we analyze a particular spatial locality case (called horizontal locality) inherent to manycore accelerator architectures employing barrel execution of SPMD kernels, such as GPUs. We then propose an adaptive memory access granularity framework to exploit and enforce the horizontal locality in order to reduce the interferences among accelerator cores memory accesses and hence improve DRAM efficiency. With the proposed technique, DRAM efficiency grows by 1.42X on average, resulting in 12.3% overall performance gain, for a set of representative memory intensive GPGPU applications.
INDEX TERMS
SIMD processors, Multi-core/single-chip multiprocessors, Memory hierarchy
CITATION
Chunyang Gou, Georgi Gaydadjiev, "Exploiting SPMD Horizontal Locality", IEEE Computer Architecture Letters, vol.10, no. 1, pp. 20-23, January-June 2011, doi:10.1109/L-CA.2011.5
14 ms
(Ver 2.0)

Marketing Automation Platform Marketing Automation Tool