There is significant interest in using hardware accelerators for data management. Most of the work has focused on GPUs, but there is interest in the use of FPGAs as well. In parallel, there is research on disaggregated architectures. Initially the focus was on disaggregated storage, but now disaggregated memory is being investigated. My interest is combining both of these lines to see if/how a disaggregated and heterogeneous platform can be developed to support modern data management needs and workloads. In this talk, my objective is to present what I am thinking in hopes that we can have a discussion on its feasibility.