most researchers make up their own architectures to demonstrate place-and-route algorithms, for example there is this challenge to improve p&r: http://www.eecg.toronto.edu/~vaughn/challenge/challenge.html