Semiconductor Spin Qubits

First electrons to full high‑fidelity characterization

The promise of spin qubit systems relies crucially on the ability to mitigate crosstalk in gates, drifting potentials, and low‑level waveforms that are difficult to stabilize. Quantum Machines’ Orchestration Platform shortens the entire path from first contact to randomized benchmarking with stable, calibrated, high-fidelity qubit operations.

Whether the qubit encoding is Loss-DiVincenzo with synthetic spin–orbit fields, singlet-triplet qubits, or exchange‑only qubits, the combination of OPX1000 pulse control and QDAC-II ultralow noise DC delivery enables fully synchronized gate operations, readout, and analog feedback from a single deterministic timing engine. Early characterization steps such as leakage‑matrix measurements, stability‑diagram acquisition, and identification of single‑electron regimes are dramatically accelerated, with QDAC‑II’s native current measurements.

Once the device is operating in its qubit regime, the hardware natively supports both Pauli spin blockade and Elzerman readout, with all sequencing, alignment, timing waits, and data acquisition executed inside QM’s unique Pulse Processing Unit. OPX instructions are evaluated in real time allowing long duration experiments such as seconds‑long T₁ measurements with limited overhead, and reducing drift susceptibility. For coherent control, the OPX1000 generates high-purity ESR tones, shaped exchange pulses, and dynamically varying detuning ramps, making multi-parameter spectroscopy, Rabi, and 2D Chevron scans easy and fast. Entire maps can be captured before averaging, significantly suppressing slow environmental drift and improving extraction of resonance and coherence parameters. OPX1000 and the intuitive QUA language allow for randomized‑benchmarking sequences, going from system, bringing up to qubit control and fidelity optimization to be performed in a matter of 48 hours.

Read more

From Qubits to QPU, with QUA-enabled real-time control 

QUA, QM’s pulse-processing language, enables deterministic DC control, fast pulse sequencing, and realtime readout within a single program, without repeated client–hardware interaction in spin qubit experiments. For example, chargestabilitydiagram acquisition can be implemented by coordinating QDACII voltage stepping with OPX1000 generated trigger pulses and reflectometry readout. Nested loops in QUA define the scan geometry, while functions such as play(), wait(), and measure() synchronize voltage updates and data acquisition.  

Constructs such as loops, conditionals, arithmetic operations, and switch statements allow experiments to branch dynamically based on program variables or measurement outcomes. QUA and OPX1000 allow for closedloop routines such as PID stabilization of chargesensor, simplify the construction of complex voltagepulse sequences, and enable highlevel abstractions such as the VoltageGateSequence() class to define pulse trajectories directly in gatevoltage space. Builtin compensation ensures zero net DC offset, while dynamic crosstalk correction can be applied through configurationlevel crosstalk matrices, mitigating unintended perturbations of neighboring qubits in dense arrays. 

Ultimately, by embedding sequencing, feedback, and data handling directly into the execution layer, QUA reduces experimental overhead and allows the user to focus on physics rather than control.  

Request a Demo

Firmware for large scale spin qubit QPU control

The OPX firmware provides the signal-integrity backbone required for low-noise, stable, distortion-free spin qubit control across complex pulse sequences.

Digital filtering is implemented directly in firmware, with native FIR and IIR filters available for any pulse element. IIR filters compensate distortions from bias-tees, while FIR filters support sub-nanosecond baseband rise times and skew control at picosecond-level resolution. Because filtering runs on the Pulse Processing Unit, pulse predistortion is applied in real time, reducing memory use and minimizing experiment run time.

The firmware also manages line-shape correction and synchronized timing across DC, RF, and readout channels. Crosstalk compensation matrices can be embedded in the configuration file, enabling real-time correction of capacitive coupling between adjacent gates. For rapid, multidimensional voltage sequences, the system can automatically generate compensation pulses to maintain zero net DC offset.

The same real-time engine further supports stabilization routines such as onboard PID loops for drift-sensitive charge sensors, with update rates of a few hundred nanoseconds. Error signals, thresholds, and correction voltages are computed in firmware, allowing experiments to adapt continuously to quasi-static charge rearrangements without interruption.

Read more