Producing large numbers of cells efficiently is a prerequisite to delivering affordable cultivated meat. Existing bioreactor designs used in biopharma have so far proven inadequate. Numerous innovative designs have been proposed to overcome their limitations. Repeated prototyping is needed to evaluate and optimize these designs. However, physical prototyping is expensive and time-consuming. In biopharma, virtual prototyping using computational modeling has offered a useful alternative, providing design insight by predicting the three key drivers in bioreactor efficacy: suspension, mass transfer, and energy dissipation. The modeling methodologies are effective when producing physically robust cell lines such as CHO and bacterial cells at low densities. Under those conditions, representing the media as a fluid, and cells and aggregates as sparsely distributed rigid particles experiencing advection and differential forces, captures the properties relevant to predicting the key drivers. At higher densities, and for the more delicate cell lines required for cultivated meat production, the assumptions underlying these representations break down. The fluid becomes more slurry-like. The cells may interact physically and through chemical signaling not just with those within their own aggregate, but also with those in other aggregates. A new modeling methodology that integrates biology with fluids and mechanics is required. Our goal is to develop such a whole-system modeling methodology serving those who are innovating new bioreactor designs.