Understanding the evolution of planetary atmospheres is central to planetary science and comparative planetology. From the tenuous exospheres of Mercury and the moons of Jupiter to the dense atmosphere of Titan and Venus, atmospheric processes govern surface-space interactions, habitability potential, and the long-term stability of volatile reservoirs. These processes are inherently multi-scale, shaped by collisions, chemistry, radiation, and plasma interactions that couple microscopic particle physics with macroscopic atmospheric dynamics. Capturing this complexity requires computational methods that extend beyond traditional fluid-based models. I have developed a state-of-the-art molecular kinetics model based on the Direct Simulation Monte Carlo (DSMC) method. DSMC is a particle-based technique that represents a gas as an ensemble of computational particles and tracks their interactions to recover macroscopic properties. Unlike fluid models, DSMC can accurately simulate non-equilibrium regimes, rarefied gases, and transitions between collisional and collisionless states. These capabilities make it particularly well-suited for studying upper planetary atmospheres, where processes such as escape, photodissociation, sputtering, and ion-neutral interactions dominate. The model has been optimized for parallel computing, allowing efficient scaling to large particle numbers and complex multi-species atmospheres. The proposed project aims to leverage Dardel to perform large-scale simulations of atmospheric evolution across a diverse set of planetary environments. Specific targets include Mercury, Titan, the icy moons of Jupiter, and Earth-like exoplanets. Running these simulations requires access to a high-performance computing system capable of handling the computational demands of DSMC. The method’s reliance on large numbers of computational particles, combined with the need to resolve long physical timescales and multi-species interactions, makes it highly resource-intensive. Dardel’s architecture, with its strong parallelization support and advanced interconnect, is ideally suited for this purpose. Access to Dardel will allow me to efficiently execute and expand these simulations, producing results that are not feasible on local or smaller-scale computing systems. The broader scientific goals of this project are twofold. First, to advance our understanding of the physical mechanisms that shape atmospheric escape and long-term evolution in planetary systems. Second, to establish predictive frameworks that connect model outputs with spacecraft and telescope observations. This will strengthen collaborations with past, present, and future missions such as Cassini, Juno, and JUICE. The results will not only provide new constraints on planetary atmospheres in our Solar System, but also inform models of habitability and atmospheric stability for terrestrial exoplanets. By combining advanced particle-based modeling with the computational power of Dardel, this project will contribute to answering fundamental questions about how planetary atmospheres form, evolve, and interact with their environments.