We discuss an implementation of molecular dynamics (MD) simulations on a graphic processing unit (GPU) in the NVIDIA CUDA language. We tested our code on a modern GPU, the NVIDIA GeForce 8800 GTX. Results for two MD algorithms suitable for short-rang
ed and long-ranged interactions, and a congruential shift random number generator are presented. The performance of the GPUs is compared to their main processor counterpart. We achieve speedups of up to 80, 40 and 150 fold, respectively. With newest generation of GPUs one can run standard MD simulations at 10^7 flops/$.
We report extensive simulations of the relaxation dynamics of a self-avoiding polymer confined inside a cylindrical pore. In particular, we concentrate on examining how confinement influences the scaling behavior of the global relaxation time of the
chain, t, with the chain length N and pore diameter D. An earlier scaling analysis based on the de Gennes blob picture led to t ~ N^2D^(1/3). Our numerical effort that combines molecular dynamics and Monte Carlo simulations, however, consistently produces different t-results for N up to 2000. We argue that the previous scaling prediction is only asymptotically valid in the limit N >> D^(5/3) >> 1, which is currently inaccessible to computer simulations and, more interestingly, is also difficult to reach in experiments. Our results are thus relevant for the interpretation of recent experiments with DNA in nano- and micro-channels.
