Abstract
In cellular automata, a single update rule is applied simultaneously to each cell on the lattice. Therefore, many approaches with parallel and distributed systems have been researched. In this paper, we propose a cellular automata system which consists of only one desktop computer and one PCI board with one FPGA. In the system, users need to write only the update rule using a dedicated language (a simple extension of C), and then, the rule is compiled and linked with the hardware libraries based on a new computation method for the systems with limited memory bandwidth. In the computation method, cells on large size lattice can be efficiently processed in parallel and in pipeline. Experiments on lattice-gas model (2048 × 1024 cells) and Life Game (2048 × 1024) showed good results. The speedup by the system is 155 times and 250 times compared with Pentium-III 700MHz respectively.
Sign-in/Register to access full text options 
Free) 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
