Instruction-Set Architecture for Programmable NV-Center Quantum Repeater Nodes
Paper details
Section titled “Paper details”| Field | Value |
|---|---|
| Journal | Open MIND |
| Published | 2026-02-16 |
| Authors | Vinay Kumar, Claudio Cicconetti, Riccardo Bassoli, Marco Conti, Andrea Passarella |
| DOI | 10.48550/arxiv.2602.14995 |
The proposed ISA lets a classical controller program an NV-center repeater node through two modes: deterministic selection or coherent selection of electron-spin operations. Its instructions map to nuclear-register preparation and microwave or radio-frequency pulse sequences. The work is a formal architecture, not an experimental device result.
Architecture and instruction format
Section titled “Architecture and instruction format”The network contains (M) nodes controlled by one classical controller. At each time step (t), the controller broadcasts an instruction vector (I(t)). Each element specifies one node’s operation for that round.
A local decoder converts that instruction into register preparation, microwave pulses, and radio-frequency pulses. Microwave pulses manipulate the electron spin, while radio-frequency pulses drive nuclear-spin transitions. The electron spin (E_0) is the data qubit, and (n) nuclear spins form its control register. The electron spin is optically interfaced for entanglement distribution.
The nuclear register exposes (2^n) computational addresses. Each instruction contains OPCODE, PARAMS, PATTERN, and MODE. OPCODE names a primitive such as (X), (Z), (H), (R_y(\theta)), CNOT, or MEASURE. PARAMS carries angles, phases, and control and target identifiers. PATTERN identifies the nuclear configurations that enable the operation. MODE selects deterministic or coherent register control.
Instruction rounds are synchronized across all nodes. The shared slot should fit within the coherence time of the shortest participating qubit. Each round resets the nuclear register, applies the selected operation, and performs optional readout. The ISA separates protocol instructions from local MW and RF hardware.
The cited NV-center platform supports room-temperature operation, but the proposal does not fix a device temperature.
Deterministic and coherent register control
Section titled “Deterministic and coherent register control”Deterministic mode prepares one nuclear basis state from PATTERN. That state selects one electron-spin operation without interference between branches. For two nuclear spins, 00 selects (I), 01 selects (X), 10 selects (R_y(\theta)), and 11 selects CNOT from (E_0) to (N_A).
Coherent mode prepares the nuclear register in a superposition over the same configurations. The resulting operation is
[ U_{\mathrm{repeater}}=\sum_i c_i|a_i\rangle\langle a_i|N\otimes U{E0}(a_i). ]
Measurement of the nuclear register in a rotated basis selects a conditional electron operation. This mode can compare several operations before the node commits to one. Here, the nuclear register is a quantum address register rather than classical program memory.
Deterministic mode is the default because it maximizes repetition rate and avoids discarded outcomes. Coherent mode is intended for selective diagnostics and calibration.
BBPSSW protocol example
Section titled “BBPSSW protocol example”Two entangled electron-spin pairs at each node share a noisy channel. One pair is kept, and the other is measured after bilateral gates. The controller broadcasts deterministic instructions to both nodes. The minimal instruction set is {PAULI, CNOT, MEASURE}.
| PATTERN | OPCODE | Function |
|---|---|---|
| 000 | (I) | Idle |
| 001 | (X) | Pauli bit flip |
| 010 | (Y) | Pauli bit and phase flip |
| 011 | (Z) | Pauli phase flip |
| 100 | CNOT | Two-qubit entangling gate |
| 101 | MEASURE | Computational-basis measurement |
The protocol applies bilateral twirling, bilateral CNOT, local measurement, and a parity check. During twirling, both nodes apply the same random Pauli rotation to each pair. During the CNOT stage, the keep qubit controls the measure qubit at each node.
Both nodes measure their target qubits and exchange the classical outcomes. They keep the pairs only when the outcomes agree. A classical register can retain each measurement outcome for later parity checks.
The example does not require coherent register control. Other protocols may need opcodes for explicit memory moves or multi-qubit measurements. Measurement extends the execution model from unitary gates to CPTP maps.
Interferometric diagnostics
Section titled “Interferometric diagnostics”Coherent control provides an interference measurement that deterministic control cannot provide. One nuclear qubit is prepared in an equal superposition of two program branches. (U_0) and (U_1) act on the same electron input state.
Nuclear readout in the (X) basis creates interference between the two outputs. The probability contains (\operatorname{Re}\langle\psi|U_0^\dagger U_1|\psi\rangle), which compares the implemented branches. That overlap is inaccessible when the register selects a single branch.
Preparing a relative phase (\phi) and sweeping it extracts the real and imaginary parts of the overlap. Two settings, (\phi_1=0) and (\phi_2=-\pi/2) radians, suffice for this extraction. Each setting yields an unnormalized electron state and its outcome probability.
The measured overlap gives the state fidelity
[ F_\psi=\left|\langle\psi|U_0^\dagger U_1|\psi\rangle\right|^2. ]
This result supports fidelity witnessing and in situ pulse calibration on the probed state. It does not recover the average global gate fidelity, which the paper says requires full process tomography. Coherent control therefore gives partial channel information without reconstructing complete channels.
The method can tune a pulse sequence so the implemented operation better matches its intended unitary. The nuclear superposition coefficients define an LCU, while nuclear measurement gives heralded Kraus maps on the electron spin.
Scaling and throughput
Section titled “Scaling and throughput”The architecture generalizes from one electron spin to (E_e) electron spins per node. Each electron has its own local nuclear register (N_{e,1}) through (N_{e,r}). An independent control assumption lets (E_e) electrons execute in parallel. The same instruction format applies to every electron address.
Register size (r) sets both the instruction address space and reset burden. The modeled per-electron throughput is
[ R=\frac{1}{t_{\mathrm{MW}}+t_{\mathrm{RF}}+t_{\mathrm{meas}}+t_{\mathrm{class}}+r\tau_{\mathrm{reset}}} ]
in rounds per second. The denominator includes microwave time, radio-frequency time, measurement time, classical latency, and (r) reset operations.
The model evaluates (r) in {1, 2, 4, 8, and 16} with the other overheads fixed. Throughput falls hyperbolically as the per-nuclear-spin reset time increases. Each additional nuclear spin makes that degradation steeper under sequential reset.
The model assumes sequential reset for this trend and treats other pulse times as protocol-dependent constants. Coherent control adds a small fixed overhead for phase preparation and rotated-basis readout. The authors absorb this overhead into the combined MW, RF, and measurement time.
The condition (2^r\geq p) ensures unique register states for (p) protocol operations. Thus, (r) should be no larger than needed to cover the required instruction set. Parallel electron throughput scales ideally by (E_e), subject to controller and crosstalk constraints.
Experimental context and unresolved limits
Section titled “Experimental context and unresolved limits”The paper cites heralded entanglement over 3 metres and a Bell-inequality violation over 1.3 kilometres. These results establish the platform context, but they do not validate the proposed ISA. The proposal does not report new reach or memory-duration values.
A cited ten-qubit register combined one electron with nine 13C nuclei and provided quantum memory up to one minute. Another cited result controlled dozens of 13C spins around one NV center. Current diamond platforms have operated up to two NV electron spins within one crystal. For that reason, the paper treats multi-electron nodes as a forward-looking extension.
The throughput model is intentionally minimal and does not model crosstalk, detailed pulse sequences, or decoherence directly. The coherent diagnostic is state dependent and does not provide global gate fidelity. Coherent execution introduces postselection overhead in the form of discarded outcomes.
The synchronization assumption requires all node operations to finish before the shortest relevant coherence time expires. The paper specifies no sample geometry, defect density, isotope enrichment, or optical grade. Those specifications remain open engineering requirements for an experimental implementation.
The work includes no measured throughput, gate performance, or end-to-end network test. A practical evaluation would need those data before comparing the ISA with fixed-control repeater implementations.
Materials and support
Section titled “Materials and support”The paper’s material requirement is diamond for an NV-center quantum repeater node with optically addressed electron spins and controllable nuclear spins. 6CCVD’s diamond materials catalog lists MPCVD-grown single-crystal plates and wafers up to 125 mm. The paper gives no defect density, isotope enrichment, or optical grade, so researchers should use the linked engineering support service to define sample requirements before fabrication.
View Original Abstract
Programmability is increasingly central in emerging quantum network software stacks, yet the node-internal controller-to-hardware interface for quantum repeater devices remains under-specified. We introduce the idea of an instruction-set architecture (ISA) for controller-driven programmability of nitrogen-vacancy (NV) center quantum repeater nodes. Each node consists of an optically interfaced electron spin acting as a data qubit and a long-lived nuclear-spin register acting as a control program. We formalize two modes of programmability: (i) deterministic register control, where the nuclear register is initialized in a basis state to select a specific operation on the data qubit; and (ii) coherent register control, where the register is prepared in superposition, enabling coherent combinations of operations beyond classical programmability. Network protocols are expressed as controller-issued instruction vectors, which we illustrate through a compact realization of the BBPSSW purification protocol. We further show that coherent register control enables interferometric diagnostics such as fidelity witnessing and calibration, providing tools unavailable in classical programmability. Finally, we discuss scalability to multi-electron and multi-nuclear spin architectures and connection to Linear combination of unitaries (LCU) and Kraus formulation.
High Concentration Quantum Grade Single Crystal Diamond
View Specs →