Quantum computers are extremely difficult to scale up due to issues with noise, cooling, control, testing and manufacturing. These become more difficult as the number of qubits increases. The most reasonable way to scale up beyond a few thousand qubits (at best) is to use a modular architecture made up of multiple small quantum processing units (QPUs) that can exchange quantum information. Quantum interconnects are the physical devices used to establish an inter-QPU quantum communication channel by entangling pairs of qubits on different QPUs. While interconnects provide a clear path to scalability they come with their own set of challenges. First and foremost is the relatively slow rate of information exchange between QPUs. This is a result of low entanglement distribution rates and circuit operation overheads. In this talk we present new methods for dealing with the interconnect bottleneck at the logical level. Our results show that the minimal requirements for interconnects that can be used in near-term multi-QPU devices are not far from the current state-of-the-art.