Although large radio-telescope arrays allow us to observe the celestial sphere with an unprecedented level of detail and sensitivity, additional antennas drastically increases the cost of processing and storing their data, complicating the design of computing hardware. Our overall goal is to provide a system, which we term SimSDP, to aid in their design. It will achieve this by providing resource usage estimations for some given imaging pipeline and hardware architecture, allowing for more informed decisions when building the production systems. We lay the groundworks in this paper by presenting and validating an initial system that implements three different imaging pipelines. We find that in most cases, our system is able to accurately estimate the scaling across both algorithmic parameters as well as parallelization when compared to measured data, with errors roughly in the 1–5% range, demonstrating its ability to inform design decisions.