We describe the implementation of RipQP, an interior-point algorithm for convex quadratic optimization. Our Julia implementation is open source, and accommodates computations in multiple floating-point systems. In particular, it can be initialized in a low-precision system, such as Float32, as a form of warm start, and gradually transition through higher-precision systems until it reaches the prescribed accuracy. On platforms with hardware for various floating-point systems, our strategy results in savings in terms of time, number of normalized iterations, and energy expended during the solve. When we only dispose of double-precision hardware and we want to solve problems in a higher floating-point system such as quadruple precision, RipQP can perform some operations in double precision, while still maintaining satisfying stopping criteria.