Sharing Measurement Results with LINQ-based Feedback#

In this tutorial, we demonstrate LINQ-based feedback of data from the demodulation path of a readout sequencer, which allows sequencers to exchange acquisition data with low latency.

We cover two data types:

I/Q Data — The (weighted) integrated result triggered by the acquire or acquire_weighted Q1ASM instruction.
Thresholded Bits (TB) — a single bit indicating whether the I/Q value exceeded a threshold line in the I/Q plane.

More information can be found in the LINQ-based feedback documentation.

ID ranges:

0: disable sending
1–15: self-cast (same sequencer)
16–255: intra-cast / multi-cast (other sequencers on same or different modules)

Hardware Requirements:

A Qblox Cluster with one QRM connected in loopback mode (Output 1 → Input 1).

Setup#

First, we import the required packages and connect to the instrument.

[1]:

import matplotlib.pyplot as plt  # noqa: I001 - Ruff will otherwise put below and it won't be included in generation.
import numpy as np  # noqa: I001 - Ruff will otherwise put below and it won't be included in generation.
from __future__ import annotations

from qcodes.instrument import find_or_create_instrument

from qblox_instruments import Cluster, ClusterType

Scan For Clusters#

We scan for the available devices connected via ethernet using the Plug & Play functionality of the Qblox Instruments package (see Plug & Play for more info).

!qblox-pnp list

[2]:

cluster_ip = "10.10.200.42"
cluster_name = "cluster0"

Connect to Cluster#

We now make a connection with the Cluster.

[3]:

cluster: Cluster = find_or_create_instrument(
    Cluster,
    recreate=True,
    name=cluster_name,
    identifier=cluster_ip,
    dummy_cfg=(
        {
            2: ClusterType.CLUSTER_QCM,
            4: ClusterType.CLUSTER_QRM,
            6: ClusterType.CLUSTER_QCM_RF,
            8: ClusterType.CLUSTER_QRM_RF,
            10: ClusterType.CLUSTER_QTM,
            12: ClusterType.CLUSTER_QRC,
            16: ClusterType.CLUSTER_QSM,
        }
        if cluster_ip is None
        else None
    ),
)

cluster.reset()
print(cluster.get_system_status())

Status: ERROR, Flags: FEEDBACK_NETWORK_CALIBRATION_FAILED, Slot flags: NONE

Get connected modules#

[4]:

# QRM baseband modules
modules = cluster.get_connected_modules(lambda mod: mod.is_qrm_type and not mod.is_rf_type)
# This uses the module of the correct type with the lowest slot index
module = list(modules.values())[0]

[5]:

module.disconnect_outputs()
module.disconnect_inputs()
module.sequencer0.nco_prop_delay_comp_en(True)

module.sequencer0.mod_en_awg(True)
module.sequencer0.demod_en_acq(True)

module.sequencer0.connect_sequencer("io0_1")

Calibrating the Threshold#

Cable length, LO and NCO frequency rotate the integrated IQ result away from the real axis, shifting the classifier boundary. Calibration measures and corrects for this offset; skipping it means the data bits in your write-combine payload will be unreliable.

For details on thresholded measurements, see the rotation and thresholding documentation.

[6]:

def calibrate_threshold(tof: int, pulse_length: int, active_sequencers: list) -> None:
    """Calibrate threshold and rotation for each sequencer."""
    for sequencer in module.sequencers:
        sequencer.sync_en(False)
    prog = f"""
        set_awg_offs 15000, 0
        upd_param    {tof - 4}
        set_awg_offs 0, 0
        acquire      0, 0, {pulse_length}
        stop
    """
    for sequencer in active_sequencers:
        sequencer.sequence(
            {
                "waveforms": {},
                "program": prog,
                "acquisitions": {"single": {"index": 0, "num_bins": 1}},
                "weights": {},
            }
        )
        sequencer.arm_sequencer()
    module.start_sequencer()
    for sequencer in active_sequencers:
        acq = sequencer.get_acquisitions()["single"]["acquisition"]["bins"]["integration"]
        result = acq["path0"][0] + 1j * acq["path1"][0]
        rotation = np.mod(-np.angle(result), 2 * np.pi)
        threshold = (np.exp(1j * rotation) * result).real / 2
        if np.isnan(threshold):
            print(
                f"{sequencer.name}: calibration returned NaN — "
                "check loopback connection and pulse amplitude."
            )
        else:
            sequencer.thresholded_acq_threshold(threshold)
            sequencer.thresholded_acq_rotation(rotation * 360 / (2 * np.pi))
            print(
                f"{sequencer.name}: rotation = {np.degrees(rotation):.2f}°, threshold = {threshold:.4f}"
            )

Experiment Parameters#

[7]:

PULSE_LENGTH = 100
TIME_OF_FLIGHT = 149
ID = 16
calibrate_threshold(TIME_OF_FLIGHT, PULSE_LENGTH, [module.sequencer0])

cluster0_module4_sequencer0: rotation = 315.96°, threshold = 10.1035

What Do These Values Mean?#

The raw LINQ value is a large integer whose relationship to physical voltage depends on the module and its configuration. To understand this we sweep the AWG offset across its full range (−32768 to +32767) representing (-1V to 1V) for a QRM. and record the LINQ values at each step.

We also compare against the integrated acquisition returned by get_acquisitions. This is the same underlying measurement, but the hardware accumulates and integrates across the integration window.

Both should be perfectly linear in the input voltage — confirming that the LINQ value can be used directly as a proxy for voltage in feedback decisions.

Run Calibration Sweep#

Rather than re-uploading the sequence for each amplitude step, all 32 steps are embedded in a single Q1ASM program. The sender sweeps through 32 DC offsets with set_awg_offs, storing each acquisition in its own bin. The receiver pops each pair of LINQ payloads into dedicated registers — I into R0–R31, Q into R32–R63 — so the entire sweep completes in a single hardware run.

[21]:

amplitudes = np.linspace(-1.0, 1.0, 32)
awg_ints = np.linspace(-32768, 32767, 32).astype(int)

# Sender: configure IQ sharing once, then iterate through all 32 offsets
p_send = f"    fb_acq_iq_id    {ID}, 4\n    wait_sync       4\n"
for i, awg_int in enumerate(awg_ints):
    p_send += (
        f"    set_awg_offs    {awg_int}, 0\n"
        f"    upd_param       {TIME_OF_FLIGHT}\n"
        f"    acquire         0, {i}, {PULSE_LENGTH}\n"
    )
p_send += "    stop\n"

# Receiver: pop I into R0–R31 and Q into R32–R63 for each sender step
p_recv = "    wait_sync       4\n"
for i in range(32):
    p_recv += (
        f"    wait            600\n"
        f"    fb_pop_data     {ID}, R{i}\n"
        f"    fb_pop_data     {ID}, R{i + 32}\n"
    )
p_recv += "    stop\n"
linq_vals_i, intg_vals = [], []
module.sequencer0.sequence(
    {
        "waveforms": {},
        "program": p_send,
        "acquisitions": {"single": {"index": 0, "num_bins": 32}},
        "weights": {},
    }
)
module.sequencer1.sequence(
    {
        "waveforms": {},
        "program": p_recv,
        "acquisitions": {},
        "weights": {},
    }
)

module.arm_sequencer(0)
module.arm_sequencer(1)
module.start_sequencer()

_intg = module.sequencer0.get_acquisitions()["single"]["acquisition"]["bins"]["integration"]
for i in range(32):
    intg_vals.append(_intg["path0"][i])
    linq_vals_i.append(module.sequencer1.get_register(f"R{i}"))

Plot Results#

[22]:

def _fit(vals: list | np.ndarray) -> tuple[np.poly1d, np.ndarray]:
    """Return a degree-1 polynomial fit and its coefficients array."""
    c = np.polyfit(amplitudes, vals, 1)  # always length deg+1=2, no leading-zero stripping
    return np.poly1d(c), c


def _plot_series(ax: plt.Axes, vals: list | np.ndarray, color: str, label: str) -> np.poly1d:
    """Scatter-plot vals and overlay its linear fit on ax."""
    fit, c = _fit(vals)
    ax.scatter(amplitudes, vals, color=color, zorder=5, s=50, label=f"{label} (measured)")
    ax.plot(
        amplitudes,
        fit(amplitudes),
        color=color,
        linestyle="--",
        linewidth=1.5,
        label=f"{label} fit: {c[0]:.3g}·A + {c[1]:.3g}",
    )
    return fit

[23]:

BLUE = "#3A7DC9"
fig, (ax1, ax2) = plt.subplots(1, 2, figsize=(13, 4))

for ax, vals, title, ylabel in [
    (ax1, linq_vals_i, "LINQ fb_pop_data — I", "I value (raw LINQ units)"),
    (ax2, intg_vals, "Hardware integration path", "I value (integrated units)"),
]:
    ax.axhline(0, color="#888", linewidth=0.8, linestyle=":")
    ax.axvline(0, color="#888", linewidth=0.8, linestyle=":")
    _plot_series(ax, vals, BLUE, "I")
    ax.set_xlabel("Waveform Amplitude (normalized)")
    ax.set_ylabel(ylabel)
    ax.set_title(title)
    ax.legend(fontsize=9)
    ax.grid(True, alpha=0.25)

fig.suptitle(
    "Waveform Amplitude vs. Measured I Value — LINQ vs. Integrated", fontweight="bold", y=1.01
)
plt.tight_layout()
plt.show()

../../../../_images/products_qblox_instruments_tutorials_QRM_422_linq_acquisition_51_0.png

Stop#

Finally, let’s stop the sequencers if they haven’t already and close the instrument connection. One can also display a detailed snapshot containing the instrument parameters before closing the connection by uncommenting the corresponding lines.

[24]:

# Stop all sequencers.
module.stop_sequencer()

# Print status of sequencers 0 and 1 (should now say it is stopped).
print(module.get_sequencer_status(0))
print(module.get_sequencer_status(1))
print()

Status: ERROR, State: STOPPED, Exit Code: 0, Info Flags: FORCED_STOP, ACQ_SCOPE_DONE_PATH_0, ACQ_SCOPE_DONE_PATH_1, ACQ_BINNING_DONE, Warning Flags: NONE, Error Flags: SEQUENCE_PROCESSOR_DATA_OVERFLOW, Log: []
Status: OKAY, State: STOPPED, Exit Code: 0, Info Flags: FORCED_STOP, Warning Flags: NONE, Error Flags: NONE, Log: []

Sharing Measurement Results with LINQ-based Feedback#

Setup#

Scan For Clusters#

Connect to Cluster#

Get connected modules#

Calibrating the Threshold#

Experiment Parameters#

Sharing Thresholded Bits (TB) with Intra-cast#

Configure Module#

Experiment Parameters#

Define sender Q1ASM Program#

Define receiver Q1ASM Program#

Upload Sequences#

Run the Experiment#

Verify Result#

Sharing Raw I/Q Values with Intra-cast#

Experiment Parameters#

Define Q1ASM Programs#

Define receiver Q1ASM Program#

Upload Sequences#

Run the Experiment#

Verify Result#

What Do These Values Mean?#

Run Calibration Sweep#

Plot Results#

Stop#