As a route towards a cluster of interconnected processors in a quantum network, the Quantum Device Lab’s approach to short-range modularity uses small modules with high fabrication yield flip-chip bonded to a common carrier chip. In the first part of this talk, we focus on the hardware realization of this 3D-integrated architecture with indium bump bonds, inter-chip spacing control, and parameter targeting enabling high-fidelity operations [1,2]. We then leverage this technology in a two-module processor, with three qubits per node [3]. The first module acts as a server generating cluster states as entangled quantum resource. The second module acts as a client, consuming the resource through real-time adaptive measurement basis choice. We demonstrate that the client can implement universal single- and two-qubit gates with local measurements and rotations only. As an example of blind quantum computation, we show results of a measurement-based Deutsch-Jozsa algorithm. We verify that the computation remains private, that is, the server’s state doesn’t reveal the client’s algorithm nor its result. This demonstrates that cloud quantum providers can be set up in a way that they respect the data privacy of their clients [4].

[1] Norris et al., EPJ Quant. Tech. 11, 5 (2024)
[2] Norris et al., EPJ Quant. Tech. 13, 29 (2026)
[3] Dalton et al., PRX Quantum 6, 040365 (2025)
[4] Song et al., arXiv:2605.14656 (2026)