Demystifying the 16 × 16 thread-block for stencils on the GPU.

Siham Tabik Maurice Peemen Nicolás Guil Henk Corporaal

Published in: Concurr. Comput. Pract. Exp. (2015)

Keyphrases

real time
graphics hardware
parallel computing
heterogeneous computing
dct coefficients
image blocks
parallel computation
database
data sets
digital images
parallel processing
graphics processing units
block matching
gpu implementation
graphics processors
finer granularity