Over the course of the last decade, the Nice model has dramatically changed our view of the solar system's formation and early evolution. Within the context of this model, a transient period of planet–planet scattering is triggered by gravitational interactions between the giant planets and a massive primordial planetesimal disk, leading to a successful reproduction of the solar system's present-day architecture. In typical realizations of the Nice model, self-gravity of the planetesimal disk is routinely neglected, as it poses a computational bottleneck to the calculations. Recent analyses have shown, however, that a self-gravitating disk can exhibit behavior that is dynamically distinct, and this disparity may have significant implications for the solar system's evolutionary path. In this work, we explore this discrepancy utilizing a large suite of Nice model simulations with and without a self-gravitating planetesimal disk, taking advantage of the inherently parallel nature of graphic processing units. Our simulations demonstrate that self-consistent modeling of particle interactions does not lead to significantly different final planetary orbits from those obtained within conventional simulations. Moreover, self-gravitating calculations show similar planetesimal evolution to non-self-gravitating numerical experiments after dynamical instability is triggered, suggesting that the orbital clustering observed in the distant Kuiper Belt is unlikely to have a self-gravitational origin.