0.13.0

This release adds support for CUDA 13 and Python 3.13 and removes support for CUDA 11 and Python 3.10.
It adds support for using the CUDA-Q QEC libraries for real-time decoding on Quantinuum backends, and adds support for submission to QCI backends. Check out the release notes below to learn about additional new content.

What's Changed

Features and Enhancements 🎉

• QCI integration by @qci-opensource in #3112
• Support resource estimation on remote simulators by @atgeller in #3360
• Phase Folding Optimization by @atgeller in #3306
• Support CUDA 13 by @mitchdz in #3502
• C++ support for returning vectors of scalar types with cudaq.run by @khalatepradnya in #3514
• Python support for returning lists of simple types from cudaq.run by @1tnguyen in #3555
• Support gpu_decoder_config for Quantinuum by @1tnguyen in #3580

Bug Fixes 🐛

• Misc. bug fixes and improvements by @schweitzpgi in #3214
• Adding an error on invalid target attribute by @sacpis in #3244
• Add support for passing const references by @schweitzpgi in #3300
• Fix for issue 3239 by @schweitzpgi in #3309
• Fix failure in evolve for collapse operators by @1tnguyen in #3327
• Fix incorrect operand type conversion for some arithmetic operations in Python by @khalatepradnya in #3346
• Add missing reset_exec_ctx call in evolve_async by @1tnguyen in #3387
• Update BraketServerHelper to use ccnot instead of ccx by @jacofeld in #3418
• Fix docs typos and compiler pipeline for OQC backends by @khalatepradnya in #3441
• Automatic measurement for sample on remote targets by @khalatepradnya in #3422
• Bug fix to allow select on qubits by @bettinaheim in #3443
• Fix bug in unitary mixture calculation when the matrix values are small due to low error rate by @1tnguyen in #3465
• Fix memory leak in Array allocation by @1tnguyen in #3478
• Fixing bugs in dynamics operator conversion for super-operators by @1tnguyen in #3487
• registerDeviceKernel to also include function call declarations by @1tnguyen in #3501
• Various Python bridge fixes and improvements by @bettinaheim in #3489
• Fixes bugs in Python argument conversion by @schweitzpgi in #3536
• Various fixes for vector support in Python by @1tnguyen in #3533
• Various fixes for dynamic kernel loading by @1tnguyen in #3507
• Fix a bug in Python run, run_async, sample_async, and observe_async: callable args are dropped by @1tnguyen in #3545
• Fix bugs in apply specialization pass by @schweitzpgi in #3583

Breaking Changes 🛠

• Update the quantinuum target to use Nexus service by @khalatepradnya in #3286
• Update Docker base image to Ubuntu 24.04 by @mitchdz in #3421
• Update to cuquantum 25.09 by @1tnguyen in #3436
• Add simulator and replace runtime call for resource counting by @atgeller in #3051
• Make Python sample API spec compliant by @khalatepradnya in #3452

Documentation Updates ✏️

• Docs and examples for QCI backends by @qci-opensource in #3373
• Add Quantum Machines documentation by @khalatepradnya in #3301
• New SKQD tutorial by @zohimchandani in #3439
• Add QSCI notebook by @ikkoham in #3500
• New tutorial: QM/MM implementation with PE framework by @marwafar in #3228
• Update codes for quantum chemistry examples by @marwafar in #3303
• Grouped dynamics examples by @mawolf2023 in #2900
• Consistent docs for hardware targets by @khalatepradnya in #3339
• Update gate fusion doc by @1tnguyen in #3344
• Extend the Hamiltonian generation function to add UHF by @marwafar in #3386
• Docs edits to fix ambiguities in code by @mmvandieren in #3383
• Fix kron order used in initial state construction in example notebook by @1tnguyen in #3574

Other Changes

• Bump minimum Python version to 3.11 by @bmhowe23 in #3588
• Ability to send on the QuantinuumConfig.simulator field by @1tnguyen in #3451
• Handle additional target arguments for QCI by @khalatepradnya in #3511
• Add support for additional cuTensorNet 2.9 knobs including deterministic pathfinding/contraction by @cuMikeBrown in #3559
• Adding hyperparameters for gradient based algorithms by @sacpis in #3218
• Expose control qubits in pybind exposure of noise_model::get_channels by @wavefunction91 in #2940
• Adding a decomposition pattern for controlled-Y gate (CY) by @khalatepradnya in #3296
• Ability to inject extra payload to REST requests by @nvidia-dobri in #3292
• Support architecture-specific customization in target config by @1tnguyen in #3390
• Support apply_noise in builder mode by @1tnguyen in #3399
• Restructuring of compiler pipelines by @schweitzpgi in #3423
• Ignore unsuccessful shot in output log by @khalatepradnya in #3463
• Add new argument for maximum number of qubits for Quantinuum targets by @khalatepradnya in #3566
• Adding reset after measure if qubit is to be reused by @1tnguyen in #3450

New Contributors

• @MIWdlB made their first contribution in #3272
• @wavefunction91 made their first contribution in #2940
• @qci-opensource made their first contribution in #3112
• @hrushikesh890 made their first contribution in #3361
• @vdFLunN5FWuz made their first contribution in #3381
• @jacofeld made their first contribution in #3418
• @Renaud-K made their first contribution in #3529

Release created by workflow 19330157076.
Full Changelog: 0.12.0...0.13.0