we need to specifically target the register file port allocation and the bus multiplexing to meet the performance requirements. we cannot go all-out because the power consumption and wiring would be far too high. see: https://groups.google.com/d/msg/comp.arch/qeMsE7UxvlI/HjCDBp5yBgAJ http://lists.mailinglist.openpowerfoundation.org/pipermail/openpower-hdl-cores/2020-April/000032.html