Perovskite-silicon tandem solar cells have shown a rapid progress within the past 5 years in terms of their research cell efficiency and are currently being investigated as candidates for the next generation of industrial PV devices. This raises the question of which silicon bottom cell will be most suitable for tandem application. Currently, the silicon heterojunction (SHJ) technology dominates in tandem research achieving world records. However, it is an open issue of how to transfer these research results to industrial mass production, which is driven by cost reduction and resource efficiency and includes challenges like upscaling and long-term stability. Therefore, it is highly relevant for the PV industry to get reliable and predictive estimates on the efficiency and cost potential, as well as technologically feasible solutions. In this work, we elaborate on silicon bottom cell concepts based on the PERC, TOPCon, and SHJ technology combined with two different interconnection concepts. For each tandem device, the efficiency potential is investigated by means of an experimentally validated simulation model. Second, we evaluate the bottom cell concepts in terms of all-in cell costs per piece. Bringing the efficiency potential and cost evaluation together allows us to assess the different tandem cell concepts in terms of all-in module cost per watt peak. Our results show that perovskite-silicon tandem devices are promising candidates to significantly reduce the levelized cost of electricity and, in particular, that the "race" for the best silicon bottom cell is still open to all the investigated bottom cell technologies.