
Parallel GPU-accelerated metaheuristics 203
User
MO
ParadisEO
GPU
ParadisEO
Software
Hardware
GPU
and neighborhood
Evaluation
Specific data
<<device>>
<<host>>
CUDA
(1)
(2)
(3)
(1) Allocate and copy of data
(2) Parallel neighborhood evaluation
(3) Copy of evaluation result
CPU
Predefined neighbor
FIGURE 9.3. The skeleton of the ParadisEO-MO-GPU.
and particle swarm optimization. Nevertheless, the library is designed in such
a way to allow further extensions for other metaheuristics and it is still in
development phase by the authors. The parallelization strategy adopted on
GPU is principally based on fitness evaluation. The sequential structure of
the optimization algorithms ...