v0.3.0

Changelog

• 8051294 Bump to v0.3.0
• 532c264 docs: finish RFC 0005 — the reference matches the graph that shipped
• 1289693 feat(qpi-driver): make the calibration graph run on a chip, not just the simulator
• 0d14824 docs(rfc): RFC 0005 is implemented
• 4d4c89d feat(qpi-driver): parametrise the coupler-driven CZ, completing RFC 0005
• fe74672 feat(qpi-driver): measure the coupler's parking current
• d0cb1d9 feat(qpi-driver): give the coupler a frequency of its own
• e49242a feat(qpi-driver): close the three open questions in RFC 0005
• 888b9f4 fix(qpi-driver): keep a coupler drive's frequency through a held offset
• 18837e3 feat(qpi-driver): find the frequency a parametric CZ must be driven at
• 4cdc7df feat(qpi-driver): cancel what the EF pulse leaks into the ground state
• ee6109a feat(qpi-driver): refine f12 with a Ramsey on the 1-2 transition
• a26bab4 feat(qpi-driver): drive the EF transition on the ladder, in one frame
• acb85a7 feat(qpi-driver): check the dispersive ladder is evenly spaced
• d2c80a3 feat(qpi-driver): refine the EF pi pulse by amplifying its error
• d91dfa8 feat(qpi-driver): measure leakage into the third level
• 31c1505 feat(qpi-driver): the simulated acquisition reports every level, not two
• 1882fff docs(rfc): an EF pulse has no envelope, so drag_12 cannot be written yet
• ad6e387 docs(rfc): three-state readout is blocked in the simulator, not the routines
• 4a5590b feat(qpi-driver): measure the pi pulse on the 1-2 transition
• 1b9d9d6 feat(qpi-driver): discriminated readout gets its own operating point
• c90b015 fix(qpi-driver): each qubit gets discriminated against its own line
• 0e6eea5 feat(qpi-driver): measure the dispersive shift the readout rests on
• 534ff43 docs(rfc): phase 2 is done, all three of it
• c3be24f feat(qpi-driver): spectroscopy chooses its own drive power
• fa0758d docs: your chip's config is not part of this codebase
• d42a621 feat(qpi-driver): drive the |1>-|2> transition, and measure where it is
• 73b206a feat(qpi-driver): a dispersive readout, and the discriminator it needs calibrating
• 8bdd6f3 feat(qpi-driver): measure the readout chain's timing instead of hand-setting it
• 3c16b3a docs: this is not a lab's project
• 3f3cea2 feat(qpi-driver): measure whether a calibration is stale instead of assuming
• ecda1c2 docs(rfc): RFC 0005 — complete the calibration graph
• b5b87d0 docs(rfc): put the simulator's new claims where its limits are stated
• fa453ae feat(qpi-driver): simulate the readout chain and the pulse envelope
• b756e40 test(qpi-driver): run the whole DAG through the real stack, not a fake
• b8bce43 fix(qpi-driver): the CZ chevron could not write what it measured
• d1ffb65 docs(rfc): record what qblox parity found, and what is still assumed
• f1db050 docs(changelog): say what changed, not how it was diagnosed
• 18ecb07 feat(qpi-driver): a coupler can declare the gap its CZ bridges
• 9463eb0 feat(qpi-ui): a failed job says why
• 2d5c85d test(qpi-driver): the whole loop runs under both schedulers
• e913e6f feat(qpi-driver): is_simulated works on qblox too
• 4340bf7 fix(scripts): a stub CLI reported every documented flag as removed
• 94be3c5 fix(qpi-driver): the DRAG sweep was eleven orders of magnitude out
• fa89830 feat(qpi-driver): raw traces over more than one qubit
• c1fe43a feat(qpi-driver): park couplers over SPI or a QCM, chosen by config
• fc0c60e test(e2e): assert what each measurement level actually renders
• e217142 feat(qpi-driver): the coupler CZ is a parametric sideband
• f2ed9fd fix: the coupler edge, its parking current, and an empty IQ plot
• 9ca7238 feat(e2e): point the dashboard's job console at the simulated chip
• 1c17071 fix(qpi-driver): the three acquisition protocols return three shapes
• d9543dc fix(qpi-driver): a virtual Z did nothing at all
• 7a6f66c docs(changelog): record two-qubit gates and the three fits they corrected
• 02f8937 feat(qpi-driver): a flux pulse is a CZ now, not a refusal
• 4bbd403 feat(qpi-driver): a coupled pair, so a CZ is a thing that can be wrong
• a901dfb docs(rfc): settle the read rule on calibration reports
• bdd156d fix(qpi-driver): refuse a spectroscopy line the sweep never resolved
• 620023e feat(qpi-driver): -o is_simulated=true, for a node with no hardware
• 4ce092b feat(qpi-driver): a simulated chip to put where the cluster goes
• 70319e2 fix(qpi-driver): close() raised, leaving the cluster held
• 553a69b test(qpi-driver): calibrate end to end against the simulator
• b6f9bb8 test(qpi-driver): add the scqubits physics simulator (RFC 0004 §7, tier 3)
• 365a53c docs: document calibrate usage, and correct what was wrong
• 1be8910 docs: fix the tuners page's links and its Makefile claim
• af2c6bc docs(rfc): record the dashboard panel as built
• 2997251 feat(dashboard): add the calibration panel (RFC 0004 §6.8)
• 7bed76a docs(rfc): bring RFC 0004 in line with what was built
• 9ce9efd feat(qpi-ui): give calibrate a dispatch path and the SDKs the operation
• 699a7d0 feat(qpi-driver): implement the calibration tuners for real
• 27f0a94 Merge feature/calibration-driver into the corrected RFC branch
• b3dbc69 docs(rfc): add RFC 0004 (calibration tuners), corrected against the code
• 29e5584 feat: Add calibrate operation support and quantify/qblox tuners