Obviously, at our end, we would need quite an excellent classical computer to transfer our ideas into single qubits to be sent to the remote node. Luckily we are already quite good at this. The most crucial part is that the small and easy to access nodes just need as much as one qubit of quantum computational power. Could you tell why only one qubit would be enough?
Only one qubit is needed because blind quantum computing only requires an entangled pair and no real local computation. Thus, only one qubit is being sent.