This research program will use first-principles radiative plasma simulations to understand how neutron stars radiate. Neutron stars are the culprits of the most infamous astrophysical emission enigmas: 1) pulsar radio emission, 2) multi-messenger signals of compact-object binary mergers, and 3) simultaneous generation of giant flares and fast radio bursts from magnetars. These emission mechanisms have remained elusive because we do not have a self-consistent theory that combines plasma physics (describing microscopic motions and energy dissipation of the magnetized gas) and radiative processes (describing the reprocessing of the energy into radiation). This project combines the forefront plasma physics theory with exascale high-performance computing technologies to achieve two breakthroughs: 1) generation of first-principles 3D models of the radiative plasmas around pulsars, mergers, and magnetars; and 2) development of a novel open-source simulation toolkit for self-consistent and high-fidelity modeling of astroplasmas. These enable a quantitative understanding of the unsolved emission mechanisms (including efficiency, variability, and output spectra) and direct comparison to observations. Analyzing astronomical observations with these superior physics-constrained models enable direct tests of their validity and a leap in improving the accuracy of the modern nuclear/particle physics theories of the still-unknown neutron star equation of state. The PI has a world-leading role in computational astroplasma physics, an established record of impactful and innovative research in the astrophysics of neutron stars, and 10 years of experience in state-of-the-art high-performance computing solutions.