Comment by storus

I think the op meant pipeline parallelism where during inference you only transfer the activation between layers where you cut the model in two, which shouldn't be too large.