| Popis: |
Computer Science development plays an important role to understand natural phenomena. Its advance has impacted on studies results from many areas such as Biology and Medicine. Agent-based Models (ABM) are an alternative to model and to simulate natural phenomena. Nowadays, ABM simulation of complex systems can be very limited, mostly about execution time and performance. Among all alternatives to improve performance, GPUs can be very attractive mainly because new architectures are being developed and their new features can make these systems even more precise. In this work, we present the advantages to use GPUs to model ABMs, modeling an ABM of the human immune system. We give a brief introduction to the ABM, review the GPU architectures and then we show how they can be useful to model these systems. Finally, we compare CPU x GPU implementations and discuss the results we achieved. |
